@aglyn/plugins-ai 1.0.0-beta.292 → 1.0.0-beta.293

package/src/lib/site-import/site-import-page.js.map

@@ -1 +1 @@
-{"version":3,"sources":["../../../../../../../libs/plugins/ai/src/lib/site-import/site-import-page.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\n/*==========================================\n * ONE CRAWLED PAGE, AS AN IMPORT KEEPS IT (AGL-3976).\n *\n * The crawler's page (`SiteImportPage`) holds everything it extracted; an\n * import keeps much less, and keeps it in two shapes:\n *\n * - **Listed**, for the inventory: its address, title, how it was found and\n *   how many words and photos it has. No text: an inventory of a whole site\n *   rides the job document.\n * - **Read**, for a wave: its text in order, its sections from its headings,\n *   and its photos' addresses. The text is kept as the page says it and is\n *   fenced as untrusted data only where a prompt shows it; the section names\n *   and the title travel outside the fence, so they are neutralized (no\n *   fence marker, no hidden character) and cut short. Every block of its\n *   main region rides along too (AGL-4083), so a page can be brought in word\n *   for word.\n *\n * And from the home page, what the site's header and footer are made of\n * (AGL-4083): its brand as the source draws it (`siteImportHeaderBrand`) and\n * its own facts (`siteImportSiteFacts`), for the header and footer the import\n * builds.\n *=========================================*/\n\nimport { neutralizeUntrustedText } from '@aglyn/shared-util-tools/untrusted-text'\nimport {\n  AI_SITE_IMPORT_BLOCKED_STATUSES,\n  AI_SITE_IMPORT_IMAGES_PER_PAGE,\n  AI_SITE_IMPORT_LABEL_MAX_CHARS,\n  AI_SITE_IMPORT_MAX_SECTIONS,\n  AI_SITE_IMPORT_PAGE_TEXT_MAX_CHARS,\n  AI_SITE_IMPORT_READ_MAX_CHARS,\n  aiSiteImportCapBlocks,\n  type AiSiteImportLeftOutReason,\n  type AiSiteImportListedPage,\n  type AiSiteImportPageRead,\n} from '../model/ai-site-import'\nimport type { SiteImportPage, SiteImportSkipReason } from './site-import-crawl'\nimport type { SiteImportBlock, SiteImportOrganization } from './site-import-html'\nimport { siteImportLogoCandidates, type SiteImportLogoCandidate } from './site-import-images'\n\n/** A line of untrusted text safe to show outside a fence: neutralized, one line, cut short. */\nexport function siteImportLabel(text: string, max = AI_SITE_IMPORT_LABEL_MAX_CHARS): string {\n  const line = neutralizeUntrustedText(text).replace(/\\s+/g, ' ').trim()\n  return line.length > max ? `${line.slice(0, max - 1).trimEnd()}…` : line\n}\n\nconst wordsIn = (text: string) => (text.match(/[\\p{L}\\p{N}]+/gu) ?? []).length\n\n/** How many words a page says, its title and headings included; a link inside a sentence is counted once. */\nexport function siteImportPageWords(page: Pick<SiteImportPage, 'title' | 'blocks'>): number {\n  return page.blocks.reduce(\n    (total, block) => total + (block.kind === 'image' || (block.kind === 'link' && block.inline) ? 0 : wordsIn(block.text)),\n    0,\n  )\n}\n\n/** Between a page's own name and its site's in a browser title: \"About | Rosie's Bakery\". */\nconst TITLE_SEPARATOR = /\\s+[|\\u2013\\u2014\\u00b7:-]\\s+/\n\n/**\n * A page's name on the new site: \"Home\" for the home page, else its browser\n * title without the old site's name after a separator (\"About | Rosie's\n * Bakery\" is \"About\"), else its first heading.\n */\nexport function siteImportPageTitle(page: Pick<SiteImportPage, 'path' | 'title' | 'blocks'>): string {\n  if ((page.path.replace(/\\/+$/, '') || '/') === '/') return 'Home'\n  const own = siteImportLabel((page.title ?? '').split(TITLE_SEPARATOR)[0] ?? '')\n  if (own.length >= 2) return own\n  return siteImportLabel(page.blocks.find((block) => block.kind === 'heading')?.text ?? '')\n}\n\n/** The page as the inventory lists it. */\nexport function siteImportListedPage(page: SiteImportPage): AiSiteImportListedPage {\n  return {\n    url: page.url,\n    path: page.path,\n    title: siteImportPageTitle(page),\n    foundBy: page.foundBy,\n    words: siteImportPageWords(page),\n    images: page.images.length,\n  }\n}\n\n/** The page's photos, the ones in its body first, each once. */\nexport function siteImportPageImages(page: Pick<SiteImportPage, 'blocks' | 'images'>): Array<{ url: string; alt: string }> {\n  const seen = new Set<string>()\n  const images: Array<{ url: string; alt: string }> = []\n  const add = (url: string, alt: string) => {\n    if (!url || seen.has(url) || images.length >= AI_SITE_IMPORT_IMAGES_PER_PAGE) return\n    seen.add(url)\n    images.push({ url, alt: siteImportLabel(alt, 250) })\n  }\n  for (const block of page.blocks) if (block.kind === 'image' && block.region === 'main') add(block.url, block.alt)\n  for (const image of page.images) add(image.url, image.alt)\n  return images\n}\n\n/**\n * The page's sections, from its headings: the first is its opening (its\n * title or first heading), then one for each second-level heading of its\n * body, at most `AI_SITE_IMPORT_MAX_SECTIONS`. A page with no headings below\n * its title but more to say than an opening gets a second section for it.\n */\nexport function siteImportPageSections(page: Pick<SiteImportPage, 'title' | 'blocks'>): string[] {\n  const main = page.blocks.filter((block) => block.region === 'main')\n  const headings = main.filter((block): block is Extract<typeof block, { kind: 'heading' }> => block.kind === 'heading')\n  const opening = siteImportLabel(headings.find((block) => block.level === 1)?.text || page.title || '') || 'Introduction'\n  const sections = [opening]\n  const seen = new Set([opening.toLowerCase()])\n  for (const heading of headings) {\n    if (heading.level !== 2) continue\n    const name = siteImportLabel(heading.text)\n    if (!name || seen.has(name.toLowerCase())) continue\n    seen.add(name.toLowerCase())\n    sections.push(name)\n    if (sections.length >= AI_SITE_IMPORT_MAX_SECTIONS) break\n  }\n  if (sections.length === 1 && siteImportPageWords({ title: '', blocks: main }) > 80) sections.push('Details')\n  return sections\n}\n\n/** The page's text in order, as a prompt shows it inside its fence. */\nexport function siteImportPageText(page: Pick<SiteImportPage, 'title' | 'description' | 'blocks'>): string {\n  const lines: string[] = []\n  if (page.title) lines.push(`Title: ${page.title}`)\n  if (page.description) lines.push(`Description: ${page.description}`)\n  for (const block of page.blocks) {\n    const where = block.region === 'main' ? '' : `(${block.region}) `\n    if (block.kind === 'heading') lines.push(`${'#'.repeat(block.level)} ${block.text}`)\n    else if (block.kind === 'image') lines.push(`${where}[photo: ${block.alt || 'no description'}]`)\n    else if (block.kind === 'list-item') lines.push(`${where}- ${block.text}`)\n    else if (block.kind === 'quote') lines.push(`${where}> ${block.text}`)\n    else if (block.kind === 'link') {\n      // A link inside a sentence is in its block's words already.\n      if (!block.inline) lines.push(`${where}${block.text}`)\n    } else lines.push(`${where}${block.text}`)\n  }\n  const text = lines.join('\\n')\n  return text.length > AI_SITE_IMPORT_PAGE_TEXT_MAX_CHARS ? text.slice(0, AI_SITE_IMPORT_PAGE_TEXT_MAX_CHARS) : text\n}\n\n/** The page as a wave reads it: its main region's blocks whole, within the read's budget. */\nexport function siteImportPageRead(page: SiteImportPage, at: string): AiSiteImportPageRead {\n  const main = page.blocks.filter((block) => block.region === 'main')\n  const capped = aiSiteImportCapBlocks(main, AI_SITE_IMPORT_READ_MAX_CHARS)\n  return {\n    at,\n    title: siteImportPageTitle(page),\n    text: siteImportPageText(page),\n    sections: siteImportPageSections(page),\n    images: siteImportPageImages(page),\n    blocks: capped.blocks,\n    ...(capped.cut ? { cut: true } : {}),\n    words: siteImportPageWords({ title: '', blocks: main }),\n    forms: page.forms.kept,\n  }\n}\n\n/*------------------------------------------\n * The home page's header and footer (AGL-4083)\n *-----------------------------------------*/\n\n/**\n * How the source's header draws its brand: its logo alone (aglyn.com), its\n * logo with its name beside it, or its name alone. `text` is the name as the\n * header prints it, or the site's own name when it prints nothing at all.\n */\nexport interface SiteImportHeaderBrand {\n  kind: 'logo-only' | 'logo-and-name' | 'name-only'\n  logo: SiteImportLogoCandidate | null\n  /** Whether the header prints the name as words. */\n  name: boolean\n  text: string\n}\n\n/** The site's own name: its head's, its structured data's, else its home page's title before a separator. */\nexport function siteImportSiteName(home: Pick<SiteImportPage, 'siteName' | 'organization' | 'title'>): string {\n  return siteImportLabel(home.siteName || home.organization?.name || (home.title ?? '').split(TITLE_SEPARATOR).pop() || '')\n}\n\n/** Words short enough to be a name, not a sentence. */\nconst looksLikeName = (text: string) => text.length >= 2 && text.length <= AI_SITE_IMPORT_LABEL_MAX_CHARS && text.split(/\\s+/).length <= 8\n\n/**\n * The home page's header brand (AGL-4083), as the source shows it: the link\n * home in its header, the picture inside it, and whether words sit beside\n * the picture, inside the link or next to it. A header with no picture link\n * home is the name alone.\n */\nexport function siteImportHeaderBrand(\n  home: Pick<SiteImportPage, 'brand' | 'images' | 'icons' | 'siteName' | 'organization' | 'title' | 'blocks'>,\n): SiteImportHeaderBrand {\n  const logo = siteImportLogoCandidates(home).find((candidate) => candidate.source === 'header') ?? null\n  const siteName = siteImportSiteName(home)\n  const linked = siteImportLabel(home.brand?.text ?? '')\n  const beside = home.blocks.find(\n    (block) =>\n      block.region === 'header' &&\n      (block.kind === 'paragraph' || block.kind === 'heading' || block.kind === 'link') &&\n      !!siteName &&\n      block.text.trim().toLowerCase() === siteName.toLowerCase(),\n  )\n  const printed = looksLikeName(linked) ? linked : beside && beside.kind !== 'image' ? siteImportLabel(beside.text) : ''\n  if (logo && printed) return { kind: 'logo-and-name', logo, name: true, text: printed }\n  if (logo) return { kind: 'logo-only', logo, name: false, text: '' }\n  return { kind: 'name-only', logo: null, name: true, text: printed || siteName }\n}\n\n/** The facts a site's header and footer carry, as the source writes them. */\nexport interface SiteImportSiteFacts {\n  header: SiteImportBlock[]\n  footer: SiteImportBlock[]\n  organization: SiteImportOrganization | null\n  /** Email addresses, from its structured data and its `mailto:` links. */\n  emails: string[]\n  /** Phone numbers as written, from its structured data and its `tel:` links. */\n  phones: string[]\n  /** Its profiles on other sites: its structured data's `sameAs` and its links to them. */\n  social: string[]\n}\n\nconst SOCIAL_HOSTS =\n  /(?:^|\\.)(?:twitter\\.com|x\\.com|facebook\\.com|instagram\\.com|linkedin\\.com|youtube\\.com|tiktok\\.com|github\\.com|pinterest\\.com|threads\\.net|yelp\\.com|tripadvisor\\.com|nextdoor\\.com|vimeo\\.com|bsky\\.app|mastodon\\.social)$/i\n\n/**\n * The home page's own facts (AGL-4083), word for word: its header and footer\n * blocks, its Organization structured data, and the email addresses, phone\n * numbers and profiles they hold. Nothing here is invented or reworded; a\n * fact the page does not state is not here.\n */\nexport function siteImportSiteFacts(home: Pick<SiteImportPage, 'blocks' | 'organization'>): SiteImportSiteFacts {\n  const header = home.blocks.filter((block) => block.region === 'header')\n  const footer = home.blocks.filter((block) => block.region === 'footer')\n  const organization = home.organization ?? null\n  const emails: string[] = []\n  const phones: string[] = []\n  const social: string[] = []\n  const add = (list: string[], value: string) => {\n    const clean = value.trim()\n    if (clean && !list.some((entry) => entry.toLowerCase() === clean.toLowerCase())) list.push(clean)\n  }\n  if (organization?.email) add(emails, organization.email.toLowerCase())\n  if (organization?.telephone) add(phones, organization.telephone)\n  for (const url of organization?.sameAs ?? []) add(social, url)\n  for (const block of [...header, ...footer]) {\n    if (block.kind !== 'link') continue\n    if (block.href.startsWith('mailto:')) add(emails, block.href.slice('mailto:'.length))\n    else if (block.href.startsWith('tel:')) add(phones, block.text)\n    else {\n      try {\n        if (SOCIAL_HOSTS.test(new URL(block.href).hostname)) add(social, block.href)\n      } catch {\n        // A link that is not an address is no profile.\n      }\n    }\n  }\n  return { header, footer, organization, emails, phones, social }\n}\n\n/** Why the crawler could not read a page, as an import leaves it out; `status` is the answer an `http-error` got. */\nexport function siteImportLeftOutReason(reason: SiteImportSkipReason, status?: number): AiSiteImportLeftOutReason {\n  switch (reason) {\n    case 'login-form':\n      return 'login-form'\n    case 'robots':\n      return 'robots'\n    case 'too-large':\n      return 'too-large'\n    case 'off-origin-redirect':\n      return 'off-site'\n    case 'http-error':\n      return status !== undefined && AI_SITE_IMPORT_BLOCKED_STATUSES.has(status) ? 'blocked' : 'unreadable'\n    case 'not-reached':\n      return 'not-reached'\n    default:\n      return 'unreadable'\n  }\n}\n"],"names":["neutralizeUntrustedText","AI_SITE_IMPORT_BLOCKED_STATUSES","AI_SITE_IMPORT_IMAGES_PER_PAGE","AI_SITE_IMPORT_LABEL_MAX_CHARS","AI_SITE_IMPORT_MAX_SECTIONS","AI_SITE_IMPORT_PAGE_TEXT_MAX_CHARS","AI_SITE_IMPORT_READ_MAX_CHARS","aiSiteImportCapBlocks","siteImportLogoCandidates","siteImportLabel","text","max","line","replace","trim","length","slice","trimEnd","wordsIn","match","siteImportPageWords","page","blocks","reduce","total","block","kind","inline","TITLE_SEPARATOR","siteImportPageTitle","path","own","title","split","find","siteImportListedPage","url","foundBy","words","images","siteImportPageImages","seen","Set","add","alt","has","push","region","image","siteImportPageSections","headings","main","filter","opening","level","sections","toLowerCase","heading","name","siteImportPageText","lines","description","where","repeat","join","siteImportPageRead","at","capped","cut","forms","kept","siteImportSiteName","home","siteName","organization","pop","looksLikeName","siteImportHeaderBrand","logo","candidate","source","linked","brand","beside","printed","SOCIAL_HOSTS","siteImportSiteFacts","header","footer","emails","phones","social","list","value","clean","some","entry","email","telephone","sameAs","href","startsWith","test","URL","hostname","siteImportLeftOutReason","reason","status","undefined"],"mappings":";AAAA;;;;;;;;;;;;;;;CAeC,GAED;;;;;;;;;;;;;;;;;;;;;2CAqB2C,GAE3C,SAASA,uBAAuB,QAAQ,0CAAyC;AACjF,SACEC,+BAA+B,EAC/BC,8BAA8B,EAC9BC,8BAA8B,EAC9BC,2BAA2B,EAC3BC,kCAAkC,EAClCC,6BAA6B,EAC7BC,qBAAqB,QAIhB,6BAAyB;AAGhC,SAASC,wBAAwB,QAAsC,0BAAsB;AAE7F,6FAA6F,GAC7F,OAAO,SAASC,gBAAgBC,IAAY,EAAEC,MAAMR,8BAA8B;IAChF,MAAMS,OAAOZ,wBAAwBU,MAAMG,OAAO,CAAC,QAAQ,KAAKC,IAAI;IACpE,OAAOF,KAAKG,MAAM,GAAGJ,MAAM,GAAGC,KAAKI,KAAK,CAAC,GAAGL,MAAM,GAAGM,OAAO,GAAG,CAAC,CAAC,GAAGL;AACtE;AAEA,MAAMM,UAAU,CAACR;QAAkBA;WAAD,EAACA,cAAAA,KAAKS,KAAK,CAAC,ynaAAXT,cAAiC,EAAE,EAAEK,MAAM;;AAE9E,2GAA2G,GAC3G,OAAO,SAASK,oBAAoBC,IAA8C;IAChF,OAAOA,KAAKC,MAAM,CAACC,MAAM,CACvB,CAACC,OAAOC,QAAUD,QAASC,CAAAA,MAAMC,IAAI,KAAK,WAAYD,MAAMC,IAAI,KAAK,UAAUD,MAAME,MAAM,GAAI,IAAIT,QAAQO,MAAMf,IAAI,CAAA,GACrH;AAEJ;AAEA,2FAA2F,GAC3F,MAAMkB,kBAAkB;AAExB;;;;CAIC,GACD,OAAO,SAASC,oBAAoBR,IAAuD;QAE7D,SAACA;QAENA;IAHvB,IAAI,AAACA,CAAAA,KAAKS,IAAI,CAACjB,OAAO,CAAC,QAAQ,OAAO,GAAE,MAAO,KAAK,OAAO;IAC3D,MAAMkB,MAAMtB,iBAAgB,UAAA,EAACY,cAAAA,KAAKW,KAAK,YAAVX,cAAc,IAAIY,KAAK,CAACL,gBAAgB,CAAC,EAAE,YAA5C,UAAgD;IAC5E,IAAIG,IAAIhB,MAAM,IAAI,GAAG,OAAOgB;IAC5B,OAAOtB,yBAAgBY,oBAAAA,KAAKC,MAAM,CAACY,IAAI,CAAC,CAACT,QAAUA,MAAMC,IAAI,KAAK,+BAA3CL,kBAAuDX,IAAI,mBAAI;AACxF;AAEA,wCAAwC,GACxC,OAAO,SAASyB,qBAAqBd,IAAoB;IACvD,OAAO;QACLe,KAAKf,KAAKe,GAAG;QACbN,MAAMT,KAAKS,IAAI;QACfE,OAAOH,oBAAoBR;QAC3BgB,SAAShB,KAAKgB,OAAO;QACrBC,OAAOlB,oBAAoBC;QAC3BkB,QAAQlB,KAAKkB,MAAM,CAACxB,MAAM;IAC5B;AACF;AAEA,8DAA8D,GAC9D,OAAO,SAASyB,qBAAqBnB,IAA+C;IAClF,MAAMoB,OAAO,IAAIC;IACjB,MAAMH,SAA8C,EAAE;IACtD,MAAMI,MAAM,CAACP,KAAaQ;QACxB,IAAI,CAACR,OAAOK,KAAKI,GAAG,CAACT,QAAQG,OAAOxB,MAAM,IAAIb,gCAAgC;QAC9EuC,KAAKE,GAAG,CAACP;QACTG,OAAOO,IAAI,CAAC;YAAEV;YAAKQ,KAAKnC,gBAAgBmC,KAAK;QAAK;IACpD;IACA,KAAK,MAAMnB,SAASJ,KAAKC,MAAM,CAAE,IAAIG,MAAMC,IAAI,KAAK,WAAWD,MAAMsB,MAAM,KAAK,QAAQJ,IAAIlB,MAAMW,GAAG,EAAEX,MAAMmB,GAAG;IAChH,KAAK,MAAMI,SAAS3B,KAAKkB,MAAM,CAAEI,IAAIK,MAAMZ,GAAG,EAAEY,MAAMJ,GAAG;IACzD,OAAOL;AACT;AAEA;;;;;CAKC,GACD,OAAO,SAASU,uBAAuB5B,IAA8C;QAGnD6B;IAFhC,MAAMC,OAAO9B,KAAKC,MAAM,CAAC8B,MAAM,CAAC,CAAC3B,QAAUA,MAAMsB,MAAM,KAAK;IAC5D,MAAMG,WAAWC,KAAKC,MAAM,CAAC,CAAC3B,QAA+DA,MAAMC,IAAI,KAAK;IAC5G,MAAM2B,UAAU5C,gBAAgByC,EAAAA,iBAAAA,SAAShB,IAAI,CAAC,CAACT,QAAUA,MAAM6B,KAAK,KAAK,uBAAzCJ,eAA6CxC,IAAI,KAAIW,KAAKW,KAAK,IAAI,OAAO;IAC1G,MAAMuB,WAAW;QAACF;KAAQ;IAC1B,MAAMZ,OAAO,IAAIC,IAAI;QAACW,QAAQG,WAAW;KAAG;IAC5C,KAAK,MAAMC,WAAWP,SAAU;QAC9B,IAAIO,QAAQH,KAAK,KAAK,GAAG;QACzB,MAAMI,OAAOjD,gBAAgBgD,QAAQ/C,IAAI;QACzC,IAAI,CAACgD,QAAQjB,KAAKI,GAAG,CAACa,KAAKF,WAAW,KAAK;QAC3Cf,KAAKE,GAAG,CAACe,KAAKF,WAAW;QACzBD,SAAST,IAAI,CAACY;QACd,IAAIH,SAASxC,MAAM,IAAIX,6BAA6B;IACtD;IACA,IAAImD,SAASxC,MAAM,KAAK,KAAKK,oBAAoB;QAAEY,OAAO;QAAIV,QAAQ6B;IAAK,KAAK,IAAII,SAAST,IAAI,CAAC;IAClG,OAAOS;AACT;AAEA,qEAAqE,GACrE,OAAO,SAASI,mBAAmBtC,IAA8D;IAC/F,MAAMuC,QAAkB,EAAE;IAC1B,IAAIvC,KAAKW,KAAK,EAAE4B,MAAMd,IAAI,CAAC,CAAC,OAAO,EAAEzB,KAAKW,KAAK,EAAE;IACjD,IAAIX,KAAKwC,WAAW,EAAED,MAAMd,IAAI,CAAC,CAAC,aAAa,EAAEzB,KAAKwC,WAAW,EAAE;IACnE,KAAK,MAAMpC,SAASJ,KAAKC,MAAM,CAAE;QAC/B,MAAMwC,QAAQrC,MAAMsB,MAAM,KAAK,SAAS,KAAK,CAAC,CAAC,EAAEtB,MAAMsB,MAAM,CAAC,EAAE,CAAC;QACjE,IAAItB,MAAMC,IAAI,KAAK,WAAWkC,MAAMd,IAAI,CAAC,GAAG,IAAIiB,MAAM,CAACtC,MAAM6B,KAAK,EAAE,CAAC,EAAE7B,MAAMf,IAAI,EAAE;aAC9E,IAAIe,MAAMC,IAAI,KAAK,SAASkC,MAAMd,IAAI,CAAC,GAAGgB,MAAM,QAAQ,EAAErC,MAAMmB,GAAG,IAAI,iBAAiB,CAAC,CAAC;aAC1F,IAAInB,MAAMC,IAAI,KAAK,aAAakC,MAAMd,IAAI,CAAC,GAAGgB,MAAM,EAAE,EAAErC,MAAMf,IAAI,EAAE;aACpE,IAAIe,MAAMC,IAAI,KAAK,SAASkC,MAAMd,IAAI,CAAC,GAAGgB,MAAM,EAAE,EAAErC,MAAMf,IAAI,EAAE;aAChE,IAAIe,MAAMC,IAAI,KAAK,QAAQ;YAC9B,4DAA4D;YAC5D,IAAI,CAACD,MAAME,MAAM,EAAEiC,MAAMd,IAAI,CAAC,GAAGgB,QAAQrC,MAAMf,IAAI,EAAE;QACvD,OAAOkD,MAAMd,IAAI,CAAC,GAAGgB,QAAQrC,MAAMf,IAAI,EAAE;IAC3C;IACA,MAAMA,OAAOkD,MAAMI,IAAI,CAAC;IACxB,OAAOtD,KAAKK,MAAM,GAAGV,qCAAqCK,KAAKM,KAAK,CAAC,GAAGX,sCAAsCK;AAChH;AAEA,2FAA2F,GAC3F,OAAO,SAASuD,mBAAmB5C,IAAoB,EAAE6C,EAAU;IACjE,MAAMf,OAAO9B,KAAKC,MAAM,CAAC8B,MAAM,CAAC,CAAC3B,QAAUA,MAAMsB,MAAM,KAAK;IAC5D,MAAMoB,SAAS5D,sBAAsB4C,MAAM7C;IAC3C,OAAO;QACL4D;QACAlC,OAAOH,oBAAoBR;QAC3BX,MAAMiD,mBAAmBtC;QACzBkC,UAAUN,uBAAuB5B;QACjCkB,QAAQC,qBAAqBnB;QAC7BC,QAAQ6C,OAAO7C,MAAM;OACjB6C,OAAOC,GAAG,GAAG;QAAEA,KAAK;IAAK,IAAI,CAAC;QAClC9B,OAAOlB,oBAAoB;YAAEY,OAAO;YAAIV,QAAQ6B;QAAK;QACrDkB,OAAOhD,KAAKgD,KAAK,CAACC,IAAI;;AAE1B;AAmBA,2GAA2G,GAC3G,OAAO,SAASC,mBAAmBC,IAAiE;QAC9BA;QAA5BA;IAAxC,OAAO/D,gBAAgB+D,KAAKC,QAAQ,MAAID,qBAAAA,KAAKE,YAAY,qBAAjBF,mBAAmBd,IAAI,KAAI,EAACc,cAAAA,KAAKxC,KAAK,YAAVwC,cAAc,IAAIvC,KAAK,CAACL,iBAAiB+C,GAAG,MAAM;AACxH;AAEA,qDAAqD,GACrD,MAAMC,gBAAgB,CAAClE,OAAiBA,KAAKK,MAAM,IAAI,KAAKL,KAAKK,MAAM,IAAIZ,kCAAkCO,KAAKuB,KAAK,CAAC,OAAOlB,MAAM,IAAI;AAEzI;;;;;CAKC,GACD,OAAO,SAAS8D,sBACdL,IAA2G;QAE9FhE;QAEkBgE;IAF/B,MAAMM,QAAOtE,iCAAAA,yBAAyBgE,MAAMtC,IAAI,CAAC,CAAC6C,YAAcA,UAAUC,MAAM,KAAK,qBAAxExE,iCAAqF;IAClG,MAAMiE,WAAWF,mBAAmBC;IACpC,MAAMS,SAASxE,yBAAgB+D,cAAAA,KAAKU,KAAK,qBAAVV,YAAY9D,IAAI,mBAAI;IACnD,MAAMyE,SAASX,KAAKlD,MAAM,CAACY,IAAI,CAC7B,CAACT,QACCA,MAAMsB,MAAM,KAAK,YAChBtB,CAAAA,MAAMC,IAAI,KAAK,eAAeD,MAAMC,IAAI,KAAK,aAAaD,MAAMC,IAAI,KAAK,MAAK,KAC/E,CAAC,CAAC+C,YACFhD,MAAMf,IAAI,CAACI,IAAI,GAAG0C,WAAW,OAAOiB,SAASjB,WAAW;IAE5D,MAAM4B,UAAUR,cAAcK,UAAUA,SAASE,UAAUA,OAAOzD,IAAI,KAAK,UAAUjB,gBAAgB0E,OAAOzE,IAAI,IAAI;IACpH,IAAIoE,QAAQM,SAAS,OAAO;QAAE1D,MAAM;QAAiBoD;QAAMpB,MAAM;QAAMhD,MAAM0E;IAAQ;IACrF,IAAIN,MAAM,OAAO;QAAEpD,MAAM;QAAaoD;QAAMpB,MAAM;QAAOhD,MAAM;IAAG;IAClE,OAAO;QAAEgB,MAAM;QAAaoD,MAAM;QAAMpB,MAAM;QAAMhD,MAAM0E,WAAWX;IAAS;AAChF;AAeA,MAAMY,eACJ;AAEF;;;;;CAKC,GACD,OAAO,SAASC,oBAAoBd,IAAqD;QAGlEA;IAFrB,MAAMe,SAASf,KAAKlD,MAAM,CAAC8B,MAAM,CAAC,CAAC3B,QAAUA,MAAMsB,MAAM,KAAK;IAC9D,MAAMyC,SAAShB,KAAKlD,MAAM,CAAC8B,MAAM,CAAC,CAAC3B,QAAUA,MAAMsB,MAAM,KAAK;IAC9D,MAAM2B,gBAAeF,qBAAAA,KAAKE,YAAY,YAAjBF,qBAAqB;IAC1C,MAAMiB,SAAmB,EAAE;IAC3B,MAAMC,SAAmB,EAAE;IAC3B,MAAMC,SAAmB,EAAE;IAC3B,MAAMhD,MAAM,CAACiD,MAAgBC;QAC3B,MAAMC,QAAQD,MAAM/E,IAAI;QACxB,IAAIgF,SAAS,CAACF,KAAKG,IAAI,CAAC,CAACC,QAAUA,MAAMxC,WAAW,OAAOsC,MAAMtC,WAAW,KAAKoC,KAAK9C,IAAI,CAACgD;IAC7F;IACA,IAAIpB,gCAAAA,aAAcuB,KAAK,EAAEtD,IAAI8C,QAAQf,aAAauB,KAAK,CAACzC,WAAW;IACnE,IAAIkB,gCAAAA,aAAcwB,SAAS,EAAEvD,IAAI+C,QAAQhB,aAAawB,SAAS;IAC/D,KAAK,MAAM9D,eAAOsC,gCAAAA,aAAcyB,MAAM,mBAAI,EAAE,CAAExD,IAAIgD,QAAQvD;IAC1D,KAAK,MAAMX,SAAS;WAAI8D;WAAWC;KAAO,CAAE;QAC1C,IAAI/D,MAAMC,IAAI,KAAK,QAAQ;QAC3B,IAAID,MAAM2E,IAAI,CAACC,UAAU,CAAC,YAAY1D,IAAI8C,QAAQhE,MAAM2E,IAAI,CAACpF,KAAK,CAAC,UAAUD,MAAM;aAC9E,IAAIU,MAAM2E,IAAI,CAACC,UAAU,CAAC,SAAS1D,IAAI+C,QAAQjE,MAAMf,IAAI;aACzD;YACH,IAAI;gBACF,IAAI2E,aAAaiB,IAAI,CAAC,IAAIC,IAAI9E,MAAM2E,IAAI,EAAEI,QAAQ,GAAG7D,IAAIgD,QAAQlE,MAAM2E,IAAI;YAC7E,EAAE,eAAM;YACN,+CAA+C;YACjD;QACF;IACF;IACA,OAAO;QAAEb;QAAQC;QAAQd;QAAce;QAAQC;QAAQC;IAAO;AAChE;AAEA,mHAAmH,GACnH,OAAO,SAASc,wBAAwBC,MAA4B,EAAEC,MAAe;IACnF,OAAQD;QACN,KAAK;YACH,OAAO;QACT,KAAK;YACH,OAAO;QACT,KAAK;YACH,OAAO;QACT,KAAK;YACH,OAAO;QACT,KAAK;YACH,OAAOC,WAAWC,aAAa3G,gCAAgC4C,GAAG,CAAC8D,UAAU,YAAY;QAC3F,KAAK;YACH,OAAO;QACT;YACE,OAAO;IACX;AACF"}
+{"version":3,"sources":["../../../../../../../libs/plugins/ai/src/lib/site-import/site-import-page.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\n/*==========================================\n * ONE CRAWLED PAGE, AS AN IMPORT KEEPS IT (AGL-3976).\n *\n * The crawler's page (`SiteImportPage`) holds everything it extracted; an\n * import keeps much less, and keeps it in two shapes:\n *\n * - **Listed**, for the inventory: its address, title, how it was found and\n *   how many words and photos it has. No text: an inventory of a whole site\n *   rides the job document.\n * - **Read**, for a wave: its text in order, its sections from its headings,\n *   and its photos' addresses. The text is kept as the page says it and is\n *   fenced as untrusted data only where a prompt shows it; the section names\n *   and the title travel outside the fence, so they are neutralized (no\n *   fence marker, no hidden character) and cut short. Every block of its\n *   main region rides along too (AGL-4083), so a page can be brought in word\n *   for word.\n *\n * And from the home page, what the site's header and footer are made of\n * (AGL-4083): its brand as the source draws it (`siteImportHeaderBrand`) and\n * its own facts (`siteImportSiteFacts`), for the header and footer the import\n * builds.\n *=========================================*/\n\nimport { neutralizeUntrustedText } from '@aglyn/shared-util-tools/untrusted-text'\nimport {\n  AI_SITE_IMPORT_BLOCKED_STATUSES,\n  AI_SITE_IMPORT_IMAGES_PER_PAGE,\n  AI_SITE_IMPORT_LABEL_MAX_CHARS,\n  AI_SITE_IMPORT_MAX_SECTIONS,\n  AI_SITE_IMPORT_PAGE_TEXT_MAX_CHARS,\n  AI_SITE_IMPORT_READ_MAX_CHARS,\n  aiSiteImportCapBlocks,\n  AI_SITE_IMPORT_LAYOUT_MAX_CHARS,\n  siteImportLayoutChars,\n  siteImportLayoutWithin,\n  siteImportLayoutWithoutDrawings,\n  type AiSiteImportLeftOutReason,\n  type AiSiteImportListedPage,\n  type AiSiteImportPageRead,\n} from '../model/ai-site-import'\nimport type { SiteImportPage, SiteImportSkipReason } from './site-import-crawl'\nimport type { SiteImportBlock, SiteImportOrganization } from './site-import-html'\nimport { siteImportLogoCandidates, type SiteImportLogoCandidate } from './site-import-images'\n\n/** A line of untrusted text safe to show outside a fence: neutralized, one line, cut short. */\nexport function siteImportLabel(text: string, max = AI_SITE_IMPORT_LABEL_MAX_CHARS): string {\n  const line = neutralizeUntrustedText(text).replace(/\\s+/g, ' ').trim()\n  return line.length > max ? `${line.slice(0, max - 1).trimEnd()}…` : line\n}\n\nconst wordsIn = (text: string) => (text.match(/[\\p{L}\\p{N}]+/gu) ?? []).length\n\n/** How many words a page says, its title and headings included; a link inside a sentence is counted once. */\nexport function siteImportPageWords(page: Pick<SiteImportPage, 'title' | 'blocks'>): number {\n  return page.blocks.reduce(\n    (total, block) => total + (block.kind === 'image' || (block.kind === 'link' && block.inline) ? 0 : wordsIn(block.text)),\n    0,\n  )\n}\n\n/** Between a page's own name and its site's in a browser title: \"About | Rosie's Bakery\". */\nconst TITLE_SEPARATOR = /\\s+[|\\u2013\\u2014\\u00b7:-]\\s+/\n\n/**\n * A page's name on the new site: \"Home\" for the home page, else its browser\n * title without the old site's name after a separator (\"About | Rosie's\n * Bakery\" is \"About\"), else its first heading.\n */\nexport function siteImportPageTitle(page: Pick<SiteImportPage, 'path' | 'title' | 'blocks'>): string {\n  if ((page.path.replace(/\\/+$/, '') || '/') === '/') return 'Home'\n  const own = siteImportLabel((page.title ?? '').split(TITLE_SEPARATOR)[0] ?? '')\n  if (own.length >= 2) return own\n  return siteImportLabel(page.blocks.find((block) => block.kind === 'heading')?.text ?? '')\n}\n\n/** The page as the inventory lists it. */\nexport function siteImportListedPage(page: SiteImportPage): AiSiteImportListedPage {\n  return {\n    url: page.url,\n    path: page.path,\n    title: siteImportPageTitle(page),\n    foundBy: page.foundBy,\n    words: siteImportPageWords(page),\n    images: page.images.length,\n  }\n}\n\n/** The page's photos, the ones in its body first, each once. */\nexport function siteImportPageImages(page: Pick<SiteImportPage, 'blocks' | 'images'>): Array<{ url: string; alt: string }> {\n  const seen = new Set<string>()\n  const images: Array<{ url: string; alt: string }> = []\n  const add = (url: string, alt: string) => {\n    if (!url || seen.has(url) || images.length >= AI_SITE_IMPORT_IMAGES_PER_PAGE) return\n    seen.add(url)\n    images.push({ url, alt: siteImportLabel(alt, 250) })\n  }\n  for (const block of page.blocks) if (block.kind === 'image' && block.region === 'main') add(block.url, block.alt)\n  for (const image of page.images) add(image.url, image.alt)\n  return images\n}\n\n/**\n * The page's sections, from its headings: the first is its opening (its\n * title or first heading), then one for each second-level heading of its\n * body, at most `AI_SITE_IMPORT_MAX_SECTIONS`. A page with no headings below\n * its title but more to say than an opening gets a second section for it.\n */\nexport function siteImportPageSections(page: Pick<SiteImportPage, 'title' | 'blocks'>): string[] {\n  const main = page.blocks.filter((block) => block.region === 'main')\n  const headings = main.filter((block): block is Extract<typeof block, { kind: 'heading' }> => block.kind === 'heading')\n  const opening = siteImportLabel(headings.find((block) => block.level === 1)?.text || page.title || '') || 'Introduction'\n  const sections = [opening]\n  const seen = new Set([opening.toLowerCase()])\n  for (const heading of headings) {\n    if (heading.level !== 2) continue\n    const name = siteImportLabel(heading.text)\n    if (!name || seen.has(name.toLowerCase())) continue\n    seen.add(name.toLowerCase())\n    sections.push(name)\n    if (sections.length >= AI_SITE_IMPORT_MAX_SECTIONS) break\n  }\n  if (sections.length === 1 && siteImportPageWords({ title: '', blocks: main }) > 80) sections.push('Details')\n  return sections\n}\n\n/** The page's text in order, as a prompt shows it inside its fence. */\nexport function siteImportPageText(page: Pick<SiteImportPage, 'title' | 'description' | 'blocks'>): string {\n  const lines: string[] = []\n  if (page.title) lines.push(`Title: ${page.title}`)\n  if (page.description) lines.push(`Description: ${page.description}`)\n  for (const block of page.blocks) {\n    const where = block.region === 'main' ? '' : `(${block.region}) `\n    if (block.kind === 'heading') lines.push(`${'#'.repeat(block.level)} ${block.text}`)\n    else if (block.kind === 'image') lines.push(`${where}[photo: ${block.alt || 'no description'}]`)\n    else if (block.kind === 'list-item') lines.push(`${where}- ${block.text}`)\n    else if (block.kind === 'quote') lines.push(`${where}> ${block.text}`)\n    else if (block.kind === 'link') {\n      // A link inside a sentence is in its block's words already.\n      if (!block.inline) lines.push(`${where}${block.text}`)\n    } else lines.push(`${where}${block.text}`)\n  }\n  const text = lines.join('\\n')\n  return text.length > AI_SITE_IMPORT_PAGE_TEXT_MAX_CHARS ? text.slice(0, AI_SITE_IMPORT_PAGE_TEXT_MAX_CHARS) : text\n}\n\n/**\n * The page as a wave reads it: its main region's blocks whole, within the\n * read's budget, and how the source lays them out (AGL-4106) over the blocks\n * kept, its icons' drawings dropped first when it runs past its own budget.\n */\nexport function siteImportPageRead(page: SiteImportPage, at: string): AiSiteImportPageRead {\n  const main = page.blocks.filter((block) => block.region === 'main')\n  const capped = aiSiteImportCapBlocks(main, AI_SITE_IMPORT_READ_MAX_CHARS)\n  let layout = page.layout ? siteImportLayoutWithin(page.layout, capped.blocks.length) : null\n  if (layout && siteImportLayoutChars(layout) > AI_SITE_IMPORT_LAYOUT_MAX_CHARS) layout = siteImportLayoutWithoutDrawings(layout)\n  if (layout && siteImportLayoutChars(layout) > AI_SITE_IMPORT_LAYOUT_MAX_CHARS) layout = null\n  return {\n    at,\n    title: siteImportPageTitle(page),\n    text: siteImportPageText(page),\n    sections: siteImportPageSections(page),\n    images: siteImportPageImages(page),\n    blocks: capped.blocks,\n    ...(capped.cut ? { cut: true } : {}),\n    words: siteImportPageWords({ title: '', blocks: main }),\n    forms: page.forms.kept,\n    ...(layout ? { layout } : {}),\n    // The source's own title and description, for the page's listing (AGL-4107).\n    meta: { title: siteImportMetaText(page.title, SITE_IMPORT_META_TITLE_MAX_CHARS), description: siteImportMetaText(page.description, SITE_IMPORT_META_DESCRIPTION_MAX_CHARS) },\n  }\n}\n\n/** The most characters of a source's title and description a read keeps: far past what any listing shows. */\nexport const SITE_IMPORT_META_TITLE_MAX_CHARS = 300\nexport const SITE_IMPORT_META_DESCRIPTION_MAX_CHARS = 1_000\n\n/** A title or a description as the source writes it, one line, neutralized; a longer one is kept to its ceiling. */\nfunction siteImportMetaText(text: string | null | undefined, max: number): string {\n  return neutralizeUntrustedText(String(text ?? '')).replace(/\\s+/g, ' ').trim().slice(0, max)\n}\n\n/*------------------------------------------\n * The home page's header and footer (AGL-4083)\n *-----------------------------------------*/\n\n/**\n * How the source's header draws its brand: its logo alone (aglyn.com), its\n * logo with its name beside it, or its name alone. `text` is the name as the\n * header prints it, or the site's own name when it prints nothing at all.\n */\nexport interface SiteImportHeaderBrand {\n  kind: 'logo-only' | 'logo-and-name' | 'name-only'\n  logo: SiteImportLogoCandidate | null\n  /** Whether the header prints the name as words. */\n  name: boolean\n  text: string\n}\n\n/** The site's own name: its head's, its structured data's, else its home page's title before a separator. */\nexport function siteImportSiteName(home: Pick<SiteImportPage, 'siteName' | 'organization' | 'title'>): string {\n  return siteImportLabel(home.siteName || home.organization?.name || (home.title ?? '').split(TITLE_SEPARATOR).pop() || '')\n}\n\n/** Words short enough to be a name, not a sentence. */\nconst looksLikeName = (text: string) => text.length >= 2 && text.length <= AI_SITE_IMPORT_LABEL_MAX_CHARS && text.split(/\\s+/).length <= 8\n\n/**\n * The home page's header brand (AGL-4083), as the source shows it: the link\n * home in its header, the picture inside it, and whether words sit beside\n * the picture, inside the link or next to it. A header with no picture link\n * home is the name alone.\n */\nexport function siteImportHeaderBrand(\n  home: Pick<SiteImportPage, 'brand' | 'images' | 'icons' | 'siteName' | 'organization' | 'title' | 'blocks'>,\n): SiteImportHeaderBrand {\n  const logo = siteImportLogoCandidates(home).find((candidate) => candidate.source === 'header') ?? null\n  const siteName = siteImportSiteName(home)\n  const linked = siteImportLabel(home.brand?.text ?? '')\n  const beside = home.blocks.find(\n    (block) =>\n      block.region === 'header' &&\n      (block.kind === 'paragraph' || block.kind === 'heading' || block.kind === 'link') &&\n      !!siteName &&\n      block.text.trim().toLowerCase() === siteName.toLowerCase(),\n  )\n  const printed = looksLikeName(linked) ? linked : beside && beside.kind !== 'image' ? siteImportLabel(beside.text) : ''\n  if (logo && printed) return { kind: 'logo-and-name', logo, name: true, text: printed }\n  if (logo) return { kind: 'logo-only', logo, name: false, text: '' }\n  return { kind: 'name-only', logo: null, name: true, text: printed || siteName }\n}\n\n/** The facts a site's header and footer carry, as the source writes them. */\nexport interface SiteImportSiteFacts {\n  header: SiteImportBlock[]\n  footer: SiteImportBlock[]\n  organization: SiteImportOrganization | null\n  /** Email addresses, from its structured data and its `mailto:` links. */\n  emails: string[]\n  /** Phone numbers as written, from its structured data and its `tel:` links. */\n  phones: string[]\n  /** Its profiles on other sites: its structured data's `sameAs` and its links to them. */\n  social: string[]\n}\n\nconst SOCIAL_HOSTS =\n  /(?:^|\\.)(?:twitter\\.com|x\\.com|facebook\\.com|instagram\\.com|linkedin\\.com|youtube\\.com|tiktok\\.com|github\\.com|pinterest\\.com|threads\\.net|yelp\\.com|tripadvisor\\.com|nextdoor\\.com|vimeo\\.com|bsky\\.app|mastodon\\.social)$/i\n\n/**\n * The home page's own facts (AGL-4083), word for word: its header and footer\n * blocks, its Organization structured data, and the email addresses, phone\n * numbers and profiles they hold. Nothing here is invented or reworded; a\n * fact the page does not state is not here.\n */\nexport function siteImportSiteFacts(home: Pick<SiteImportPage, 'blocks' | 'organization'>): SiteImportSiteFacts {\n  const header = home.blocks.filter((block) => block.region === 'header')\n  const footer = home.blocks.filter((block) => block.region === 'footer')\n  const organization = home.organization ?? null\n  const emails: string[] = []\n  const phones: string[] = []\n  const social: string[] = []\n  const add = (list: string[], value: string) => {\n    const clean = value.trim()\n    if (clean && !list.some((entry) => entry.toLowerCase() === clean.toLowerCase())) list.push(clean)\n  }\n  if (organization?.email) add(emails, organization.email.toLowerCase())\n  if (organization?.telephone) add(phones, organization.telephone)\n  for (const url of organization?.sameAs ?? []) add(social, url)\n  for (const block of [...header, ...footer]) {\n    if (block.kind !== 'link') continue\n    if (block.href.startsWith('mailto:')) add(emails, block.href.slice('mailto:'.length))\n    else if (block.href.startsWith('tel:')) add(phones, block.text)\n    else {\n      try {\n        if (SOCIAL_HOSTS.test(new URL(block.href).hostname)) add(social, block.href)\n      } catch {\n        // A link that is not an address is no profile.\n      }\n    }\n  }\n  return { header, footer, organization, emails, phones, social }\n}\n\n/** Why the crawler could not read a page, as an import leaves it out; `status` is the answer an `http-error` got. */\nexport function siteImportLeftOutReason(reason: SiteImportSkipReason, status?: number): AiSiteImportLeftOutReason {\n  switch (reason) {\n    case 'login-form':\n      return 'login-form'\n    case 'robots':\n      return 'robots'\n    case 'too-large':\n      return 'too-large'\n    case 'off-origin-redirect':\n      return 'off-site'\n    case 'http-error':\n      return status !== undefined && AI_SITE_IMPORT_BLOCKED_STATUSES.has(status) ? 'blocked' : 'unreadable'\n    case 'not-reached':\n      return 'not-reached'\n    case 'not-html':\n      return 'file'\n    default:\n      return 'unreadable'\n  }\n}\n"],"names":["neutralizeUntrustedText","AI_SITE_IMPORT_BLOCKED_STATUSES","AI_SITE_IMPORT_IMAGES_PER_PAGE","AI_SITE_IMPORT_LABEL_MAX_CHARS","AI_SITE_IMPORT_MAX_SECTIONS","AI_SITE_IMPORT_PAGE_TEXT_MAX_CHARS","AI_SITE_IMPORT_READ_MAX_CHARS","aiSiteImportCapBlocks","AI_SITE_IMPORT_LAYOUT_MAX_CHARS","siteImportLayoutChars","siteImportLayoutWithin","siteImportLayoutWithoutDrawings","siteImportLogoCandidates","siteImportLabel","text","max","line","replace","trim","length","slice","trimEnd","wordsIn","match","siteImportPageWords","page","blocks","reduce","total","block","kind","inline","TITLE_SEPARATOR","siteImportPageTitle","path","own","title","split","find","siteImportListedPage","url","foundBy","words","images","siteImportPageImages","seen","Set","add","alt","has","push","region","image","siteImportPageSections","headings","main","filter","opening","level","sections","toLowerCase","heading","name","siteImportPageText","lines","description","where","repeat","join","siteImportPageRead","at","capped","layout","cut","forms","kept","meta","siteImportMetaText","SITE_IMPORT_META_TITLE_MAX_CHARS","SITE_IMPORT_META_DESCRIPTION_MAX_CHARS","String","siteImportSiteName","home","siteName","organization","pop","looksLikeName","siteImportHeaderBrand","logo","candidate","source","linked","brand","beside","printed","SOCIAL_HOSTS","siteImportSiteFacts","header","footer","emails","phones","social","list","value","clean","some","entry","email","telephone","sameAs","href","startsWith","test","URL","hostname","siteImportLeftOutReason","reason","status","undefined"],"mappings":";AAAA;;;;;;;;;;;;;;;CAeC,GAED;;;;;;;;;;;;;;;;;;;;;2CAqB2C,GAE3C,SAASA,uBAAuB,QAAQ,0CAAyC;AACjF,SACEC,+BAA+B,EAC/BC,8BAA8B,EAC9BC,8BAA8B,EAC9BC,2BAA2B,EAC3BC,kCAAkC,EAClCC,6BAA6B,EAC7BC,qBAAqB,EACrBC,+BAA+B,EAC/BC,qBAAqB,EACrBC,sBAAsB,EACtBC,+BAA+B,QAI1B,6BAAyB;AAGhC,SAASC,wBAAwB,QAAsC,0BAAsB;AAE7F,6FAA6F,GAC7F,OAAO,SAASC,gBAAgBC,IAAY,EAAEC,MAAMZ,8BAA8B;IAChF,MAAMa,OAAOhB,wBAAwBc,MAAMG,OAAO,CAAC,QAAQ,KAAKC,IAAI;IACpE,OAAOF,KAAKG,MAAM,GAAGJ,MAAM,GAAGC,KAAKI,KAAK,CAAC,GAAGL,MAAM,GAAGM,OAAO,GAAG,CAAC,CAAC,GAAGL;AACtE;AAEA,MAAMM,UAAU,CAACR;QAAkBA;WAAD,EAACA,cAAAA,KAAKS,KAAK,CAAC,ynaAAXT,cAAiC,EAAE,EAAEK,MAAM;;AAE9E,2GAA2G,GAC3G,OAAO,SAASK,oBAAoBC,IAA8C;IAChF,OAAOA,KAAKC,MAAM,CAACC,MAAM,CACvB,CAACC,OAAOC,QAAUD,QAASC,CAAAA,MAAMC,IAAI,KAAK,WAAYD,MAAMC,IAAI,KAAK,UAAUD,MAAME,MAAM,GAAI,IAAIT,QAAQO,MAAMf,IAAI,CAAA,GACrH;AAEJ;AAEA,2FAA2F,GAC3F,MAAMkB,kBAAkB;AAExB;;;;CAIC,GACD,OAAO,SAASC,oBAAoBR,IAAuD;QAE7D,SAACA;QAENA;IAHvB,IAAI,AAACA,CAAAA,KAAKS,IAAI,CAACjB,OAAO,CAAC,QAAQ,OAAO,GAAE,MAAO,KAAK,OAAO;IAC3D,MAAMkB,MAAMtB,iBAAgB,UAAA,EAACY,cAAAA,KAAKW,KAAK,YAAVX,cAAc,IAAIY,KAAK,CAACL,gBAAgB,CAAC,EAAE,YAA5C,UAAgD;IAC5E,IAAIG,IAAIhB,MAAM,IAAI,GAAG,OAAOgB;IAC5B,OAAOtB,yBAAgBY,oBAAAA,KAAKC,MAAM,CAACY,IAAI,CAAC,CAACT,QAAUA,MAAMC,IAAI,KAAK,+BAA3CL,kBAAuDX,IAAI,mBAAI;AACxF;AAEA,wCAAwC,GACxC,OAAO,SAASyB,qBAAqBd,IAAoB;IACvD,OAAO;QACLe,KAAKf,KAAKe,GAAG;QACbN,MAAMT,KAAKS,IAAI;QACfE,OAAOH,oBAAoBR;QAC3BgB,SAAShB,KAAKgB,OAAO;QACrBC,OAAOlB,oBAAoBC;QAC3BkB,QAAQlB,KAAKkB,MAAM,CAACxB,MAAM;IAC5B;AACF;AAEA,8DAA8D,GAC9D,OAAO,SAASyB,qBAAqBnB,IAA+C;IAClF,MAAMoB,OAAO,IAAIC;IACjB,MAAMH,SAA8C,EAAE;IACtD,MAAMI,MAAM,CAACP,KAAaQ;QACxB,IAAI,CAACR,OAAOK,KAAKI,GAAG,CAACT,QAAQG,OAAOxB,MAAM,IAAIjB,gCAAgC;QAC9E2C,KAAKE,GAAG,CAACP;QACTG,OAAOO,IAAI,CAAC;YAAEV;YAAKQ,KAAKnC,gBAAgBmC,KAAK;QAAK;IACpD;IACA,KAAK,MAAMnB,SAASJ,KAAKC,MAAM,CAAE,IAAIG,MAAMC,IAAI,KAAK,WAAWD,MAAMsB,MAAM,KAAK,QAAQJ,IAAIlB,MAAMW,GAAG,EAAEX,MAAMmB,GAAG;IAChH,KAAK,MAAMI,SAAS3B,KAAKkB,MAAM,CAAEI,IAAIK,MAAMZ,GAAG,EAAEY,MAAMJ,GAAG;IACzD,OAAOL;AACT;AAEA;;;;;CAKC,GACD,OAAO,SAASU,uBAAuB5B,IAA8C;QAGnD6B;IAFhC,MAAMC,OAAO9B,KAAKC,MAAM,CAAC8B,MAAM,CAAC,CAAC3B,QAAUA,MAAMsB,MAAM,KAAK;IAC5D,MAAMG,WAAWC,KAAKC,MAAM,CAAC,CAAC3B,QAA+DA,MAAMC,IAAI,KAAK;IAC5G,MAAM2B,UAAU5C,gBAAgByC,EAAAA,iBAAAA,SAAShB,IAAI,CAAC,CAACT,QAAUA,MAAM6B,KAAK,KAAK,uBAAzCJ,eAA6CxC,IAAI,KAAIW,KAAKW,KAAK,IAAI,OAAO;IAC1G,MAAMuB,WAAW;QAACF;KAAQ;IAC1B,MAAMZ,OAAO,IAAIC,IAAI;QAACW,QAAQG,WAAW;KAAG;IAC5C,KAAK,MAAMC,WAAWP,SAAU;QAC9B,IAAIO,QAAQH,KAAK,KAAK,GAAG;QACzB,MAAMI,OAAOjD,gBAAgBgD,QAAQ/C,IAAI;QACzC,IAAI,CAACgD,QAAQjB,KAAKI,GAAG,CAACa,KAAKF,WAAW,KAAK;QAC3Cf,KAAKE,GAAG,CAACe,KAAKF,WAAW;QACzBD,SAAST,IAAI,CAACY;QACd,IAAIH,SAASxC,MAAM,IAAIf,6BAA6B;IACtD;IACA,IAAIuD,SAASxC,MAAM,KAAK,KAAKK,oBAAoB;QAAEY,OAAO;QAAIV,QAAQ6B;IAAK,KAAK,IAAII,SAAST,IAAI,CAAC;IAClG,OAAOS;AACT;AAEA,qEAAqE,GACrE,OAAO,SAASI,mBAAmBtC,IAA8D;IAC/F,MAAMuC,QAAkB,EAAE;IAC1B,IAAIvC,KAAKW,KAAK,EAAE4B,MAAMd,IAAI,CAAC,CAAC,OAAO,EAAEzB,KAAKW,KAAK,EAAE;IACjD,IAAIX,KAAKwC,WAAW,EAAED,MAAMd,IAAI,CAAC,CAAC,aAAa,EAAEzB,KAAKwC,WAAW,EAAE;IACnE,KAAK,MAAMpC,SAASJ,KAAKC,MAAM,CAAE;QAC/B,MAAMwC,QAAQrC,MAAMsB,MAAM,KAAK,SAAS,KAAK,CAAC,CAAC,EAAEtB,MAAMsB,MAAM,CAAC,EAAE,CAAC;QACjE,IAAItB,MAAMC,IAAI,KAAK,WAAWkC,MAAMd,IAAI,CAAC,GAAG,IAAIiB,MAAM,CAACtC,MAAM6B,KAAK,EAAE,CAAC,EAAE7B,MAAMf,IAAI,EAAE;aAC9E,IAAIe,MAAMC,IAAI,KAAK,SAASkC,MAAMd,IAAI,CAAC,GAAGgB,MAAM,QAAQ,EAAErC,MAAMmB,GAAG,IAAI,iBAAiB,CAAC,CAAC;aAC1F,IAAInB,MAAMC,IAAI,KAAK,aAAakC,MAAMd,IAAI,CAAC,GAAGgB,MAAM,EAAE,EAAErC,MAAMf,IAAI,EAAE;aACpE,IAAIe,MAAMC,IAAI,KAAK,SAASkC,MAAMd,IAAI,CAAC,GAAGgB,MAAM,EAAE,EAAErC,MAAMf,IAAI,EAAE;aAChE,IAAIe,MAAMC,IAAI,KAAK,QAAQ;YAC9B,4DAA4D;YAC5D,IAAI,CAACD,MAAME,MAAM,EAAEiC,MAAMd,IAAI,CAAC,GAAGgB,QAAQrC,MAAMf,IAAI,EAAE;QACvD,OAAOkD,MAAMd,IAAI,CAAC,GAAGgB,QAAQrC,MAAMf,IAAI,EAAE;IAC3C;IACA,MAAMA,OAAOkD,MAAMI,IAAI,CAAC;IACxB,OAAOtD,KAAKK,MAAM,GAAGd,qCAAqCS,KAAKM,KAAK,CAAC,GAAGf,sCAAsCS;AAChH;AAEA;;;;CAIC,GACD,OAAO,SAASuD,mBAAmB5C,IAAoB,EAAE6C,EAAU;IACjE,MAAMf,OAAO9B,KAAKC,MAAM,CAAC8B,MAAM,CAAC,CAAC3B,QAAUA,MAAMsB,MAAM,KAAK;IAC5D,MAAMoB,SAAShE,sBAAsBgD,MAAMjD;IAC3C,IAAIkE,SAAS/C,KAAK+C,MAAM,GAAG9D,uBAAuBe,KAAK+C,MAAM,EAAED,OAAO7C,MAAM,CAACP,MAAM,IAAI;IACvF,IAAIqD,UAAU/D,sBAAsB+D,UAAUhE,iCAAiCgE,SAAS7D,gCAAgC6D;IACxH,IAAIA,UAAU/D,sBAAsB+D,UAAUhE,iCAAiCgE,SAAS;IACxF,OAAO;QACLF;QACAlC,OAAOH,oBAAoBR;QAC3BX,MAAMiD,mBAAmBtC;QACzBkC,UAAUN,uBAAuB5B;QACjCkB,QAAQC,qBAAqBnB;QAC7BC,QAAQ6C,OAAO7C,MAAM;OACjB6C,OAAOE,GAAG,GAAG;QAAEA,KAAK;IAAK,IAAI,CAAC;QAClC/B,OAAOlB,oBAAoB;YAAEY,OAAO;YAAIV,QAAQ6B;QAAK;QACrDmB,OAAOjD,KAAKiD,KAAK,CAACC,IAAI;OAClBH,SAAS;QAAEA;IAAO,IAAI,CAAC;QAC3B,6EAA6E;QAC7EI,MAAM;YAAExC,OAAOyC,mBAAmBpD,KAAKW,KAAK,EAAE0C;YAAmCb,aAAaY,mBAAmBpD,KAAKwC,WAAW,EAAEc;QAAwC;;AAE/K;AAEA,2GAA2G,GAC3G,OAAO,MAAMD,mCAAmC,IAAG;AACnD,OAAO,MAAMC,yCAAyC,KAAK;AAE3D,kHAAkH,GAClH,SAASF,mBAAmB/D,IAA+B,EAAEC,GAAW;IACtE,OAAOf,wBAAwBgF,OAAOlE,eAAAA,OAAQ,KAAKG,OAAO,CAAC,QAAQ,KAAKC,IAAI,GAAGE,KAAK,CAAC,GAAGL;AAC1F;AAmBA,2GAA2G,GAC3G,OAAO,SAASkE,mBAAmBC,IAAiE;QAC9BA;QAA5BA;IAAxC,OAAOrE,gBAAgBqE,KAAKC,QAAQ,MAAID,qBAAAA,KAAKE,YAAY,qBAAjBF,mBAAmBpB,IAAI,KAAI,EAACoB,cAAAA,KAAK9C,KAAK,YAAV8C,cAAc,IAAI7C,KAAK,CAACL,iBAAiBqD,GAAG,MAAM;AACxH;AAEA,qDAAqD,GACrD,MAAMC,gBAAgB,CAACxE,OAAiBA,KAAKK,MAAM,IAAI,KAAKL,KAAKK,MAAM,IAAIhB,kCAAkCW,KAAKuB,KAAK,CAAC,OAAOlB,MAAM,IAAI;AAEzI;;;;;CAKC,GACD,OAAO,SAASoE,sBACdL,IAA2G;QAE9FtE;QAEkBsE;IAF/B,MAAMM,QAAO5E,iCAAAA,yBAAyBsE,MAAM5C,IAAI,CAAC,CAACmD,YAAcA,UAAUC,MAAM,KAAK,qBAAxE9E,iCAAqF;IAClG,MAAMuE,WAAWF,mBAAmBC;IACpC,MAAMS,SAAS9E,yBAAgBqE,cAAAA,KAAKU,KAAK,qBAAVV,YAAYpE,IAAI,mBAAI;IACnD,MAAM+E,SAASX,KAAKxD,MAAM,CAACY,IAAI,CAC7B,CAACT,QACCA,MAAMsB,MAAM,KAAK,YAChBtB,CAAAA,MAAMC,IAAI,KAAK,eAAeD,MAAMC,IAAI,KAAK,aAAaD,MAAMC,IAAI,KAAK,MAAK,KAC/E,CAAC,CAACqD,YACFtD,MAAMf,IAAI,CAACI,IAAI,GAAG0C,WAAW,OAAOuB,SAASvB,WAAW;IAE5D,MAAMkC,UAAUR,cAAcK,UAAUA,SAASE,UAAUA,OAAO/D,IAAI,KAAK,UAAUjB,gBAAgBgF,OAAO/E,IAAI,IAAI;IACpH,IAAI0E,QAAQM,SAAS,OAAO;QAAEhE,MAAM;QAAiB0D;QAAM1B,MAAM;QAAMhD,MAAMgF;IAAQ;IACrF,IAAIN,MAAM,OAAO;QAAE1D,MAAM;QAAa0D;QAAM1B,MAAM;QAAOhD,MAAM;IAAG;IAClE,OAAO;QAAEgB,MAAM;QAAa0D,MAAM;QAAM1B,MAAM;QAAMhD,MAAMgF,WAAWX;IAAS;AAChF;AAeA,MAAMY,eACJ;AAEF;;;;;CAKC,GACD,OAAO,SAASC,oBAAoBd,IAAqD;QAGlEA;IAFrB,MAAMe,SAASf,KAAKxD,MAAM,CAAC8B,MAAM,CAAC,CAAC3B,QAAUA,MAAMsB,MAAM,KAAK;IAC9D,MAAM+C,SAAShB,KAAKxD,MAAM,CAAC8B,MAAM,CAAC,CAAC3B,QAAUA,MAAMsB,MAAM,KAAK;IAC9D,MAAMiC,gBAAeF,qBAAAA,KAAKE,YAAY,YAAjBF,qBAAqB;IAC1C,MAAMiB,SAAmB,EAAE;IAC3B,MAAMC,SAAmB,EAAE;IAC3B,MAAMC,SAAmB,EAAE;IAC3B,MAAMtD,MAAM,CAACuD,MAAgBC;QAC3B,MAAMC,QAAQD,MAAMrF,IAAI;QACxB,IAAIsF,SAAS,CAACF,KAAKG,IAAI,CAAC,CAACC,QAAUA,MAAM9C,WAAW,OAAO4C,MAAM5C,WAAW,KAAK0C,KAAKpD,IAAI,CAACsD;IAC7F;IACA,IAAIpB,gCAAAA,aAAcuB,KAAK,EAAE5D,IAAIoD,QAAQf,aAAauB,KAAK,CAAC/C,WAAW;IACnE,IAAIwB,gCAAAA,aAAcwB,SAAS,EAAE7D,IAAIqD,QAAQhB,aAAawB,SAAS;IAC/D,KAAK,MAAMpE,eAAO4C,gCAAAA,aAAcyB,MAAM,mBAAI,EAAE,CAAE9D,IAAIsD,QAAQ7D;IAC1D,KAAK,MAAMX,SAAS;WAAIoE;WAAWC;KAAO,CAAE;QAC1C,IAAIrE,MAAMC,IAAI,KAAK,QAAQ;QAC3B,IAAID,MAAMiF,IAAI,CAACC,UAAU,CAAC,YAAYhE,IAAIoD,QAAQtE,MAAMiF,IAAI,CAAC1F,KAAK,CAAC,UAAUD,MAAM;aAC9E,IAAIU,MAAMiF,IAAI,CAACC,UAAU,CAAC,SAAShE,IAAIqD,QAAQvE,MAAMf,IAAI;aACzD;YACH,IAAI;gBACF,IAAIiF,aAAaiB,IAAI,CAAC,IAAIC,IAAIpF,MAAMiF,IAAI,EAAEI,QAAQ,GAAGnE,IAAIsD,QAAQxE,MAAMiF,IAAI;YAC7E,EAAE,eAAM;YACN,+CAA+C;YACjD;QACF;IACF;IACA,OAAO;QAAEb;QAAQC;QAAQd;QAAce;QAAQC;QAAQC;IAAO;AAChE;AAEA,mHAAmH,GACnH,OAAO,SAASc,wBAAwBC,MAA4B,EAAEC,MAAe;IACnF,OAAQD;QACN,KAAK;YACH,OAAO;QACT,KAAK;YACH,OAAO;QACT,KAAK;YACH,OAAO;QACT,KAAK;YACH,OAAO;QACT,KAAK;YACH,OAAOC,WAAWC,aAAarH,gCAAgCgD,GAAG,CAACoE,UAAU,YAAY;QAC3F,KAAK;YACH,OAAO;QACT,KAAK;YACH,OAAO;QACT;YACE,OAAO;IACX;AACF"}

package/src/lib/site-import/site-import-review.d.ts

@@ -1,40 +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 { type AiSiteImportState } from '../model/ai-site-import';
-import type { AiJobItemLedger } from '../model/ai-jobs.types';
-/** The abuse queue's `source` for a site import's row. */
-export declare const SITE_IMPORT_REVIEW_SOURCE = "site-import";
-/** One row per import job. */
-export declare function siteImportReviewId(jobId: string): string;
-/** The reference staff quote: `SI-` and the job's id. */
-export declare function siteImportReviewReference(jobId: string): string;
-export type SiteImportReviewEvent = 'read' | 'wave-started' | 'wave-finished';
-export interface SiteImportReviewInput {
-    orgId: string;
-    hostId: string;
-    jobId: string;
-    state: AiSiteImportState;
-    items: readonly AiJobItemLedger[];
-    event: SiteImportReviewEvent;
-    /** The wave the event is about; absent on `read`. */
-    wave?: number;
-    nowMs: number;
-}
-/** The row's details: what was proved, read, kept, left out and brought in, rebuilt whole each time. */
-export declare function siteImportReviewDetails(input: SiteImportReviewInput): string;
-/** Writes (or rewrites) the import's row in the abuse queue. Never throws. */
-export declare function fileSiteImportReview(firestore: FirebaseFirestore.Firestore, input: SiteImportReviewInput): Promise<string | null>;

package/src/lib/site-import/site-import-review.js

@@ -1,149 +0,0 @@
-import { _ as _extends } from "@swc/helpers/_/_extends";
-/**
- * @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.
- */ /*==========================================
- * EVERY SITE IMPORT IS LOGGED FOR TRUST & SAFETY (AGL-3976).
- *
- * A site import is a one-click copy of a website, which is exactly what a
- * phishing clone is. Proving ownership is the control; this row is how a
- * person checks the control held. Every import job files ONE row in the
- * abuse queue (`abuseReports`), in the shape the outbound screen files a
- * hold (`fileOutboundHold`): the same fields the admin page and the daily
- * review already read, `source: 'site-import'`, the proved domain as text
- * (the queue never renders a reported address as a link), and the proof.
- *
- * The row is written when the site is read and rewritten as each wave
- * starts and finishes, its `reportCount` counting the waves, so the queue
- * shows one import as one row however long it runs. It holds nothing: the
- * import goes on while the row is open. Staff look at the old site and the
- * new one; a workspace bringing in a site that is not its own is locked,
- * and the row closed with what was done.
- *
- * Never throws: a log that cannot be written must not stop a person's
- * import, and the failure is a log line staff alerts read.
- *=========================================*/ import { ABUSE_REPORT_COLLECTION, ABUSE_REPORT_MAX_DETAILS } from "@aglyn/aglyn/app-utils/abuse-report";
-import { FieldValue } from "firebase-admin/firestore";
-import { AI_SITE_IMPORT_LEFT_OUT_REASONS, aiSiteImportPagesText } from "../model/ai-site-import.js";
-/** The abuse queue's `source` for a site import's row. */ export const SITE_IMPORT_REVIEW_SOURCE = 'site-import';
-/** One row per import job. */ export function siteImportReviewId(jobId) {
-    return `site-import-${jobId}`;
-}
-/** The reference staff quote: `SI-` and the job's id. */ export function siteImportReviewReference(jobId) {
-    return `SI-${jobId.slice(0, 10).toUpperCase()}`;
-}
-const PROOF_WORDS = {
-    dns: 'a DNS TXT record',
-    meta: 'a meta tag on its home page',
-    email: 'a code emailed to an address at the domain'
-};
-/** The row's details: what was proved, read, kept, left out and brought in, rebuilt whole each time. */ export function siteImportReviewDetails(input) {
-    var _PROOF_WORDS_state_ownership_method, _state_leftOutCount_loginform;
-    const { state, items } = input;
-    const proof = state.ownership.method ? (_PROOF_WORDS_state_ownership_method = PROOF_WORDS[state.ownership.method]) != null ? _PROOF_WORDS_state_ownership_method : state.ownership.method : 'an unknown method';
-    const provenOn = state.ownership.provenAt ? new Date(state.ownership.provenAt).toISOString().slice(0, 10) : 'an unknown day';
-    const leftOut = AI_SITE_IMPORT_LEFT_OUT_REASONS.filter((reason)=>{
-        var _state_leftOutCount_reason;
-        return ((_state_leftOutCount_reason = state.leftOutCount[reason]) != null ? _state_leftOutCount_reason : 0) > 0;
-    }).map((reason)=>`${reason} ${state.leftOutCount[reason]}`).join(', ');
-    const built = items.filter((row)=>row.op === 'page' && (row.status === 'succeeded' || row.status === 'degraded')).length;
-    const started = state.waves.filter((wave)=>wave.status !== 'waiting').length;
-    const photos = state.pages.reduce((total, page)=>{
-        var _ref;
-        var _page_photos;
-        return total + ((_ref = (_page_photos = page.photos) == null ? void 0 : _page_photos.brought) != null ? _ref : 0);
-    }, 0);
-    const lines = [
-        `Site import: ${state.domain} into site ${input.hostId} (AI job ${input.jobId}).`,
-        `Ownership proven by ${proof}${state.ownership.detail ? ` (${state.ownership.detail})` : ''} on ${provenOn}.`,
-        `Read ${aiSiteImportPagesText(state.found)}; bringing in ${aiSiteImportPagesText(state.pages.length)} in ${state.waves.length} ${state.waves.length === 1 ? 'wave' : 'waves'}` + `${state.pageLimit === null ? '' : ` (plan page count ${state.pageLimit}, ${state.existingPages} already on the site)`}.`,
-        ...leftOut ? [
-            `Left out: ${leftOut}.`
-        ] : [],
-        ...((_state_leftOutCount_loginform = state.leftOutCount['login-form']) != null ? _state_leftOutCount_loginform : 0) > 0 ? [
-            `Sign-in pages refused: ${state.leftOutCount['login-form']}. A site with sign-in pages is worth a closer look.`
-        ] : [],
-        `Waves started: ${started} of ${state.waves.length}. Pages brought in so far: ${built}. Photos copied into the media library: ${photos}.`,
-        'Nothing is published by the import: its pages are drafts until a member publishes them.',
-        'Look at the old site and the new one. If this copies a site the workspace does not own, lock the workspace, then close this row with what you did; otherwise dismiss it.'
-    ];
-    return lines.join('\n').slice(0, ABUSE_REPORT_MAX_DETAILS);
-}
-/** Writes (or rewrites) the import's row in the abuse queue. Never throws. */ export async function fileSiteImportReview(firestore, input) {
-    const reviewId = siteImportReviewId(input.jobId);
-    try {
-        var _input_state_leftOutCount_loginform;
-        const ref = firestore.collection(ABUSE_REPORT_COLLECTION).doc(reviewId);
-        const first = !(await ref.get()).exists;
-        const built = input.items.filter((row)=>row.op === 'page' && (row.status === 'succeeded' || row.status === 'degraded')).length;
-        await ref.set(_extends({
-            reference: siteImportReviewReference(input.jobId),
-            // "Impersonation or trademark": the harm a copied site does is
-            // wearing a business it is not. Normal severity: the proof already
-            // held, and this row is the check that it did.
-            category: 'impersonation',
-            severity: 'normal',
-            source: SITE_IMPORT_REVIEW_SOURCE,
-            // The proved site as TEXT: the queue never renders it as a link.
-            url: `https://${input.state.domain}/`,
-            reportedHostname: input.state.domain,
-            hostId: input.hostId,
-            orgId: input.orgId,
-            details: siteImportReviewDetails(input),
-            reporterEmail: null,
-            reporterName: null,
-            dmca: null
-        }, input.event === 'wave-started' ? {
-            reportCount: FieldValue.increment(1)
-        } : first ? {
-            reportCount: 0
-        } : {}, {
-            siteImport: _extends({
-                jobId: input.jobId,
-                domain: input.state.domain,
-                method: input.state.ownership.method,
-                provenAt: input.state.ownership.provenAt,
-                detail: input.state.ownership.detail,
-                found: input.state.found,
-                kept: input.state.pages.length,
-                waves: input.state.waves.length,
-                wavesStarted: input.state.waves.filter((wave)=>wave.status !== 'waiting').length,
-                pagesBrought: built,
-                loginPagesRefused: (_input_state_leftOutCount_loginform = input.state.leftOutCount['login-form']) != null ? _input_state_leftOutCount_loginform : 0,
-                lastEvent: input.event
-            }, input.wave !== undefined ? {
-                lastWave: input.wave
-            } : {}, {
-                atMs: input.nowMs
-            }),
-            updatedAt: FieldValue.serverTimestamp()
-        }, first ? {
-            status: 'open',
-            createdAt: FieldValue.serverTimestamp()
-        } : {}), {
-            merge: true
-        });
-        return reviewId;
-    } catch (error) {
-        console.error('[site-import] trust and safety row failed', {
-            orgId: input.orgId,
-            jobId: input.jobId,
-            error
-        });
-        return null;
-    }
-}
-
-//# sourceMappingURL=site-import-review.js.map

package/src/lib/site-import/site-import-review.js.map

@@ -1 +0,0 @@
-{"version":3,"sources":["../../../../../../../libs/plugins/ai/src/lib/site-import/site-import-review.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\n/*==========================================\n * EVERY SITE IMPORT IS LOGGED FOR TRUST & SAFETY (AGL-3976).\n *\n * A site import is a one-click copy of a website, which is exactly what a\n * phishing clone is. Proving ownership is the control; this row is how a\n * person checks the control held. Every import job files ONE row in the\n * abuse queue (`abuseReports`), in the shape the outbound screen files a\n * hold (`fileOutboundHold`): the same fields the admin page and the daily\n * review already read, `source: 'site-import'`, the proved domain as text\n * (the queue never renders a reported address as a link), and the proof.\n *\n * The row is written when the site is read and rewritten as each wave\n * starts and finishes, its `reportCount` counting the waves, so the queue\n * shows one import as one row however long it runs. It holds nothing: the\n * import goes on while the row is open. Staff look at the old site and the\n * new one; a workspace bringing in a site that is not its own is locked,\n * and the row closed with what was done.\n *\n * Never throws: a log that cannot be written must not stop a person's\n * import, and the failure is a log line staff alerts read.\n *=========================================*/\n\nimport { ABUSE_REPORT_COLLECTION, ABUSE_REPORT_MAX_DETAILS } from '@aglyn/aglyn/app-utils/abuse-report'\nimport { FieldValue } from 'firebase-admin/firestore'\nimport {\n  AI_SITE_IMPORT_LEFT_OUT_REASONS,\n  aiSiteImportPagesText,\n  type AiSiteImportState,\n} from '../model/ai-site-import'\nimport type { AiJobItemLedger } from '../model/ai-jobs.types'\n\n/** The abuse queue's `source` for a site import's row. */\nexport const SITE_IMPORT_REVIEW_SOURCE = 'site-import'\n\n/** One row per import job. */\nexport function siteImportReviewId(jobId: string): string {\n  return `site-import-${jobId}`\n}\n\n/** The reference staff quote: `SI-` and the job's id. */\nexport function siteImportReviewReference(jobId: string): string {\n  return `SI-${jobId.slice(0, 10).toUpperCase()}`\n}\n\nconst PROOF_WORDS: Record<string, string> = {\n  dns: 'a DNS TXT record',\n  meta: 'a meta tag on its home page',\n  email: 'a code emailed to an address at the domain',\n}\n\nexport type SiteImportReviewEvent = 'read' | 'wave-started' | 'wave-finished'\n\nexport interface SiteImportReviewInput {\n  orgId: string\n  hostId: string\n  jobId: string\n  state: AiSiteImportState\n  items: readonly AiJobItemLedger[]\n  event: SiteImportReviewEvent\n  /** The wave the event is about; absent on `read`. */\n  wave?: number\n  nowMs: number\n}\n\n/** The row's details: what was proved, read, kept, left out and brought in, rebuilt whole each time. */\nexport function siteImportReviewDetails(input: SiteImportReviewInput): string {\n  const { state, items } = input\n  const proof = state.ownership.method ? (PROOF_WORDS[state.ownership.method] ?? state.ownership.method) : 'an unknown method'\n  const provenOn = state.ownership.provenAt ? new Date(state.ownership.provenAt).toISOString().slice(0, 10) : 'an unknown day'\n  const leftOut = AI_SITE_IMPORT_LEFT_OUT_REASONS.filter((reason) => (state.leftOutCount[reason] ?? 0) > 0)\n    .map((reason) => `${reason} ${state.leftOutCount[reason]}`)\n    .join(', ')\n  const built = items.filter((row) => row.op === 'page' && (row.status === 'succeeded' || row.status === 'degraded')).length\n  const started = state.waves.filter((wave) => wave.status !== 'waiting').length\n  const photos = state.pages.reduce((total, page) => total + (page.photos?.brought ?? 0), 0)\n  const lines = [\n    `Site import: ${state.domain} into site ${input.hostId} (AI job ${input.jobId}).`,\n    `Ownership proven by ${proof}${state.ownership.detail ? ` (${state.ownership.detail})` : ''} on ${provenOn}.`,\n    `Read ${aiSiteImportPagesText(state.found)}; bringing in ${aiSiteImportPagesText(state.pages.length)} in ${state.waves.length} ${state.waves.length === 1 ? 'wave' : 'waves'}` +\n      `${state.pageLimit === null ? '' : ` (plan page count ${state.pageLimit}, ${state.existingPages} already on the site)`}.`,\n    ...(leftOut ? [`Left out: ${leftOut}.`] : []),\n    ...((state.leftOutCount['login-form'] ?? 0) > 0\n      ? [`Sign-in pages refused: ${state.leftOutCount['login-form']}. A site with sign-in pages is worth a closer look.`]\n      : []),\n    `Waves started: ${started} of ${state.waves.length}. Pages brought in so far: ${built}. Photos copied into the media library: ${photos}.`,\n    'Nothing is published by the import: its pages are drafts until a member publishes them.',\n    'Look at the old site and the new one. If this copies a site the workspace does not own, lock the workspace, then close this row with what you did; otherwise dismiss it.',\n  ]\n  return lines.join('\\n').slice(0, ABUSE_REPORT_MAX_DETAILS)\n}\n\n/** Writes (or rewrites) the import's row in the abuse queue. Never throws. */\nexport async function fileSiteImportReview(\n  firestore: FirebaseFirestore.Firestore,\n  input: SiteImportReviewInput,\n): Promise<string | null> {\n  const reviewId = siteImportReviewId(input.jobId)\n  try {\n    const ref = firestore.collection(ABUSE_REPORT_COLLECTION).doc(reviewId)\n    const first = !(await ref.get()).exists\n    const built = input.items.filter((row) => row.op === 'page' && (row.status === 'succeeded' || row.status === 'degraded')).length\n    await ref.set(\n      {\n        reference: siteImportReviewReference(input.jobId),\n        // \"Impersonation or trademark\": the harm a copied site does is\n        // wearing a business it is not. Normal severity: the proof already\n        // held, and this row is the check that it did.\n        category: 'impersonation',\n        severity: 'normal',\n        source: SITE_IMPORT_REVIEW_SOURCE,\n        // The proved site as TEXT: the queue never renders it as a link.\n        url: `https://${input.state.domain}/`,\n        reportedHostname: input.state.domain,\n        hostId: input.hostId,\n        orgId: input.orgId,\n        details: siteImportReviewDetails(input),\n        reporterEmail: null,\n        reporterName: null,\n        dmca: null,\n        // One a wave started, so a long import reads as the volume it is.\n        ...(input.event === 'wave-started' ? { reportCount: FieldValue.increment(1) } : first ? { reportCount: 0 } : {}),\n        siteImport: {\n          jobId: input.jobId,\n          domain: input.state.domain,\n          method: input.state.ownership.method,\n          provenAt: input.state.ownership.provenAt,\n          detail: input.state.ownership.detail,\n          found: input.state.found,\n          kept: input.state.pages.length,\n          waves: input.state.waves.length,\n          wavesStarted: input.state.waves.filter((wave) => wave.status !== 'waiting').length,\n          pagesBrought: built,\n          loginPagesRefused: input.state.leftOutCount['login-form'] ?? 0,\n          lastEvent: input.event,\n          ...(input.wave !== undefined ? { lastWave: input.wave } : {}),\n          atMs: input.nowMs,\n        },\n        updatedAt: FieldValue.serverTimestamp(),\n        ...(first ? { status: 'open', createdAt: FieldValue.serverTimestamp() } : {}),\n      },\n      { merge: true },\n    )\n    return reviewId\n  } catch (error) {\n    console.error('[site-import] trust and safety row failed', { orgId: input.orgId, jobId: input.jobId, error })\n    return null\n  }\n}\n"],"names":["ABUSE_REPORT_COLLECTION","ABUSE_REPORT_MAX_DETAILS","FieldValue","AI_SITE_IMPORT_LEFT_OUT_REASONS","aiSiteImportPagesText","SITE_IMPORT_REVIEW_SOURCE","siteImportReviewId","jobId","siteImportReviewReference","slice","toUpperCase","PROOF_WORDS","dns","meta","email","siteImportReviewDetails","input","state","items","proof","ownership","method","provenOn","provenAt","Date","toISOString","leftOut","filter","reason","leftOutCount","map","join","built","row","op","status","length","started","waves","wave","photos","pages","reduce","total","page","brought","lines","domain","hostId","detail","found","pageLimit","existingPages","fileSiteImportReview","firestore","reviewId","ref","collection","doc","first","get","exists","set","reference","category","severity","source","url","reportedHostname","orgId","details","reporterEmail","reporterName","dmca","event","reportCount","increment","siteImport","kept","wavesStarted","pagesBrought","loginPagesRefused","lastEvent","undefined","lastWave","atMs","nowMs","updatedAt","serverTimestamp","createdAt","merge","error","console"],"mappings":";AAAA;;;;;;;;;;;;;;;CAeC,GAED;;;;;;;;;;;;;;;;;;;;2CAoB2C,GAE3C,SAASA,uBAAuB,EAAEC,wBAAwB,QAAQ,sCAAqC;AACvG,SAASC,UAAU,QAAQ,2BAA0B;AACrD,SACEC,+BAA+B,EAC/BC,qBAAqB,QAEhB,6BAAyB;AAGhC,wDAAwD,GACxD,OAAO,MAAMC,4BAA4B,cAAa;AAEtD,4BAA4B,GAC5B,OAAO,SAASC,mBAAmBC,KAAa;IAC9C,OAAO,CAAC,YAAY,EAAEA,OAAO;AAC/B;AAEA,uDAAuD,GACvD,OAAO,SAASC,0BAA0BD,KAAa;IACrD,OAAO,CAAC,GAAG,EAAEA,MAAME,KAAK,CAAC,GAAG,IAAIC,WAAW,IAAI;AACjD;AAEA,MAAMC,cAAsC;IAC1CC,KAAK;IACLC,MAAM;IACNC,OAAO;AACT;AAgBA,sGAAsG,GACtG,OAAO,SAASC,wBAAwBC,KAA4B;QAE1BL,qCAcjCM;IAfP,MAAM,EAAEA,KAAK,EAAEC,KAAK,EAAE,GAAGF;IACzB,MAAMG,QAAQF,MAAMG,SAAS,CAACC,MAAM,IAAIV,sCAAAA,WAAW,CAACM,MAAMG,SAAS,CAACC,MAAM,CAAC,YAAnCV,sCAAuCM,MAAMG,SAAS,CAACC,MAAM,GAAI;IACzG,MAAMC,WAAWL,MAAMG,SAAS,CAACG,QAAQ,GAAG,IAAIC,KAAKP,MAAMG,SAAS,CAACG,QAAQ,EAAEE,WAAW,GAAGhB,KAAK,CAAC,GAAG,MAAM;IAC5G,MAAMiB,UAAUvB,gCAAgCwB,MAAM,CAAC,CAACC;YAAYX;eAAD,EAACA,6BAAAA,MAAMY,YAAY,CAACD,OAAO,YAA1BX,6BAA8B,KAAK;OACpGa,GAAG,CAAC,CAACF,SAAW,GAAGA,OAAO,CAAC,EAAEX,MAAMY,YAAY,CAACD,OAAO,EAAE,EACzDG,IAAI,CAAC;IACR,MAAMC,QAAQd,MAAMS,MAAM,CAAC,CAACM,MAAQA,IAAIC,EAAE,KAAK,UAAWD,CAAAA,IAAIE,MAAM,KAAK,eAAeF,IAAIE,MAAM,KAAK,UAAS,GAAIC,MAAM;IAC1H,MAAMC,UAAUpB,MAAMqB,KAAK,CAACX,MAAM,CAAC,CAACY,OAASA,KAAKJ,MAAM,KAAK,WAAWC,MAAM;IAC9E,MAAMI,SAASvB,MAAMwB,KAAK,CAACC,MAAM,CAAC,CAACC,OAAOC;;YAAkBA;eAATD,kBAASC,eAAAA,KAAKJ,MAAM,qBAAXI,aAAaC,OAAO,mBAAI;OAAI;IACxF,MAAMC,QAAQ;QACZ,CAAC,aAAa,EAAE7B,MAAM8B,MAAM,CAAC,WAAW,EAAE/B,MAAMgC,MAAM,CAAC,SAAS,EAAEhC,MAAMT,KAAK,CAAC,EAAE,CAAC;QACjF,CAAC,oBAAoB,EAAEY,QAAQF,MAAMG,SAAS,CAAC6B,MAAM,GAAG,CAAC,EAAE,EAAEhC,MAAMG,SAAS,CAAC6B,MAAM,CAAC,CAAC,CAAC,GAAG,GAAG,IAAI,EAAE3B,SAAS,CAAC,CAAC;QAC7G,CAAC,KAAK,EAAElB,sBAAsBa,MAAMiC,KAAK,EAAE,cAAc,EAAE9C,sBAAsBa,MAAMwB,KAAK,CAACL,MAAM,EAAE,IAAI,EAAEnB,MAAMqB,KAAK,CAACF,MAAM,CAAC,CAAC,EAAEnB,MAAMqB,KAAK,CAACF,MAAM,KAAK,IAAI,SAAS,SAAS,GAC5K,GAAGnB,MAAMkC,SAAS,KAAK,OAAO,KAAK,CAAC,kBAAkB,EAAElC,MAAMkC,SAAS,CAAC,EAAE,EAAElC,MAAMmC,aAAa,CAAC,qBAAqB,CAAC,CAAC,CAAC,CAAC;WACvH1B,UAAU;YAAC,CAAC,UAAU,EAAEA,QAAQ,CAAC,CAAC;SAAC,GAAG,EAAE;WACxC,EAACT,gCAAAA,MAAMY,YAAY,CAAC,aAAa,YAAhCZ,gCAAoC,KAAK,IAC1C;YAAC,CAAC,uBAAuB,EAAEA,MAAMY,YAAY,CAAC,aAAa,CAAC,mDAAmD,CAAC;SAAC,GACjH,EAAE;QACN,CAAC,eAAe,EAAEQ,QAAQ,IAAI,EAAEpB,MAAMqB,KAAK,CAACF,MAAM,CAAC,2BAA2B,EAAEJ,MAAM,wCAAwC,EAAEQ,OAAO,CAAC,CAAC;QACzI;QACA;KACD;IACD,OAAOM,MAAMf,IAAI,CAAC,MAAMtB,KAAK,CAAC,GAAGR;AACnC;AAEA,4EAA4E,GAC5E,OAAO,eAAeoD,qBACpBC,SAAsC,EACtCtC,KAA4B;IAE5B,MAAMuC,WAAWjD,mBAAmBU,MAAMT,KAAK;IAC/C,IAAI;YAmCuBS;QAlCzB,MAAMwC,MAAMF,UAAUG,UAAU,CAACzD,yBAAyB0D,GAAG,CAACH;QAC9D,MAAMI,QAAQ,CAAC,AAAC,CAAA,MAAMH,IAAII,GAAG,EAAC,EAAGC,MAAM;QACvC,MAAM7B,QAAQhB,MAAME,KAAK,CAACS,MAAM,CAAC,CAACM,MAAQA,IAAIC,EAAE,KAAK,UAAWD,CAAAA,IAAIE,MAAM,KAAK,eAAeF,IAAIE,MAAM,KAAK,UAAS,GAAIC,MAAM;QAChI,MAAMoB,IAAIM,GAAG,CACX;YACEC,WAAWvD,0BAA0BQ,MAAMT,KAAK;YAChD,+DAA+D;YAC/D,mEAAmE;YACnE,+CAA+C;YAC/CyD,UAAU;YACVC,UAAU;YACVC,QAAQ7D;YACR,iEAAiE;YACjE8D,KAAK,CAAC,QAAQ,EAAEnD,MAAMC,KAAK,CAAC8B,MAAM,CAAC,CAAC,CAAC;YACrCqB,kBAAkBpD,MAAMC,KAAK,CAAC8B,MAAM;YACpCC,QAAQhC,MAAMgC,MAAM;YACpBqB,OAAOrD,MAAMqD,KAAK;YAClBC,SAASvD,wBAAwBC;YACjCuD,eAAe;YACfC,cAAc;YACdC,MAAM;WAEFzD,MAAM0D,KAAK,KAAK,iBAAiB;YAAEC,aAAazE,WAAW0E,SAAS,CAAC;QAAG,IAAIjB,QAAQ;YAAEgB,aAAa;QAAE,IAAI,CAAC;YAC9GE,YAAY;gBACVtE,OAAOS,MAAMT,KAAK;gBAClBwC,QAAQ/B,MAAMC,KAAK,CAAC8B,MAAM;gBAC1B1B,QAAQL,MAAMC,KAAK,CAACG,SAAS,CAACC,MAAM;gBACpCE,UAAUP,MAAMC,KAAK,CAACG,SAAS,CAACG,QAAQ;gBACxC0B,QAAQjC,MAAMC,KAAK,CAACG,SAAS,CAAC6B,MAAM;gBACpCC,OAAOlC,MAAMC,KAAK,CAACiC,KAAK;gBACxB4B,MAAM9D,MAAMC,KAAK,CAACwB,KAAK,CAACL,MAAM;gBAC9BE,OAAOtB,MAAMC,KAAK,CAACqB,KAAK,CAACF,MAAM;gBAC/B2C,cAAc/D,MAAMC,KAAK,CAACqB,KAAK,CAACX,MAAM,CAAC,CAACY,OAASA,KAAKJ,MAAM,KAAK,WAAWC,MAAM;gBAClF4C,cAAchD;gBACdiD,iBAAiB,GAAEjE,sCAAAA,MAAMC,KAAK,CAACY,YAAY,CAAC,aAAa,YAAtCb,sCAA0C;gBAC7DkE,WAAWlE,MAAM0D,KAAK;eAClB1D,MAAMuB,IAAI,KAAK4C,YAAY;gBAAEC,UAAUpE,MAAMuB,IAAI;YAAC,IAAI,CAAC;gBAC3D8C,MAAMrE,MAAMsE,KAAK;;YAEnBC,WAAWrF,WAAWsF,eAAe;WACjC7B,QAAQ;YAAExB,QAAQ;YAAQsD,WAAWvF,WAAWsF,eAAe;QAAG,IAAI,CAAC,IAE7E;YAAEE,OAAO;QAAK;QAEhB,OAAOnC;IACT,EAAE,OAAOoC,OAAO;QACdC,QAAQD,KAAK,CAAC,6CAA6C;YAAEtB,OAAOrD,MAAMqD,KAAK;YAAE9D,OAAOS,MAAMT,KAAK;YAAEoF;QAAM;QAC3G,OAAO;IACT;AACF"}