@aglyn/plugins-ai 1.0.0-beta.288 → 1.0.0-beta.290

package/src/lib/model/ai-site-import.js

@@ -39,8 +39,15 @@ import { _ as _extends } from "@swc/helpers/_/_extends";
  *   starts the next one but them.
  * - **The price.** An imported page is an AI page (`AI_MEASURED_IMPORT_CREDITS`,
  *   about 27 credits); a wave is quoted before it starts as its pages and,
- *   for the first, its header and footer.
- *=========================================*/ import { aiCreditRangeOrdered, AI_MEASURED_IMPORT_CREDITS } from "./ai-credit-estimate.js";
+ *   for the first, its header and footer. Once the step says how each page
+ *   is built, a page copied as it is costs its listing and a page Aglyn AI
+ *   arranges its arrangement, and the quote names both (AGL-4083).
+ * - **Home first (AGL-4083).** The read's verdict (`aiSiteImportReadVerdict`)
+ *   stops an import whose home page could not be read before anything is
+ *   planned, quoted or spent: a site seen without its home page is not the
+ *   site. The person is told why and checks again.
+ *=========================================*/ import { PLATFORM_BRAND_NAME } from "@aglyn/aglyn/app-utils/platform-brand";
+import { aiCreditRangeOrdered, aiCreditRangeText, AI_MEASURED_IMPORT_CREDITS } from "./ai-credit-estimate.js";
 /** The job kind. */ export const AI_SITE_IMPORT_KIND = 'import';
 /** The input naming the proved domain the job brings in. */ export const AI_SITE_IMPORT_DOMAIN_INPUT = 'domain';
 /** The most pages one wave brings in. */ export const AI_SITE_IMPORT_WAVE_PAGES = 10;
@@ -51,12 +58,34 @@ import { _ as _extends } from "@swc/helpers/_/_extends";
  */ export const AI_SITE_IMPORT_MAX_PAGES = 200;
 /** The fewest pages under one address prefix that make it a collection. */ export const AI_SITE_IMPORT_COLLECTION_MIN = 3;
 /** The most photos brought in from one page. */ export const AI_SITE_IMPORT_IMAGES_PER_PAGE = 8;
-/** The most characters of one page's text a wave keeps for its model answer. */ export const AI_SITE_IMPORT_PAGE_TEXT_MAX_CHARS = 6000;
-/** The most sections one imported page is planned with. */ export const AI_SITE_IMPORT_MAX_SECTIONS = 6;
+/**
+ * The most characters of one page's text a wave keeps for the page step's
+ * prompt, which reads no more than this; every word is kept on the read's
+ * `blocks` (AGL-4083).
+ */ export const AI_SITE_IMPORT_PAGE_TEXT_MAX_CHARS = 6000;
+/** The most sections one imported page is planned with: one for every second-level heading (AGL-4083). */ export const AI_SITE_IMPORT_MAX_SECTIONS = 24;
+/** The most blocks of one page a wave's read keeps. */ export const AI_SITE_IMPORT_READ_MAX_BLOCKS = 2000;
+/** The most characters of one page's blocks a wave's read keeps: a long policy is 40,000. */ export const AI_SITE_IMPORT_READ_MAX_CHARS = 120000;
+/**
+ * The most characters of blocks the reads of one import hold at once: they
+ * ride the job document, which Firestore caps at 1 MiB, beside the
+ * inventory and the ledger. A wave's later page past it keeps what fits and
+ * says it was cut.
+ */ export const AI_SITE_IMPORT_WAVE_READ_MAX_CHARS = 400000;
+/**
+ * A page with more words of its own than this is copied as it is, never
+ * arranged (AGL-4083): a policy, an article, a long story.
+ */ export const AI_SITE_IMPORT_PROSE_WORDS = 900;
+/** The answers bot protection and firewalls turn a crawler away with: asked to sign in, refused, or told to slow down. */ export const AI_SITE_IMPORT_BLOCKED_STATUSES = new Set([
+    401,
+    403,
+    429
+]);
 /** The most characters a section's name, a page's title or a label keeps. */ export const AI_SITE_IMPORT_LABEL_MAX_CHARS = 60;
 /** A page with fewer words than this and no photo has nothing to bring in. */ export const AI_SITE_IMPORT_MIN_WORDS = 5;
 /** The item ledger slot of the site's header and footer. */ export const AI_SITE_IMPORT_FRAME_SLOT = 'frame';
 /** The ceiling one imported page reserves: its answer pass and its listing pass, at 50 each. */ export const AI_SITE_IMPORT_PAGE_CEILING_CREDITS = 100;
+/** The ceiling a page copied as it is reserves: its listing pass alone. */ export const AI_SITE_IMPORT_COPIED_CEILING_CREDITS = 50;
 /** The ceiling the header and footer reserve: one answer at 50. */ export const AI_SITE_IMPORT_FRAME_CEILING_CREDITS = 50;
 /** The most left-out pages listed by address; the rest are counted. */ export const AI_SITE_IMPORT_LEFT_OUT_LISTED = 60;
 export const AI_SITE_IMPORT_LEFT_OUT_REASONS = [
@@ -67,8 +96,171 @@ export const AI_SITE_IMPORT_LEFT_OUT_REASONS = [
     'unreadable',
     'blocked',
     'too-large',
-    'off-site'
+    'off-site',
+    'not-reached'
+];
+/** The reasons that mean the page was there and AglynBot could not read it: the "could not be read" of what was found. */ export const AI_SITE_IMPORT_UNREAD_REASONS = [
+    'unreadable',
+    'blocked',
+    'too-large',
+    'not-reached'
 ];
+const pathOfUrl = (url)=>{
+    try {
+        return new URL(url).pathname.replace(/\/+$/, '') || '/';
+    } catch (unused) {
+        return '';
+    }
+};
+const wordsOf = (text)=>{
+    var _text_match;
+    return ((_text_match = text.match(RegExp("[\\u0041-\\u005A\\u0061-\\u007A\\u00AA\\u00B5\\u00BA\\u00C0-\\u00D6\\u00D8-\\u00F6\\u00F8-\\u02C1\\u02C6-\\u02D1\\u02E0-\\u02E4\\u02EC\\u02EE\\u0370-\\u0374\\u0376-\\u0377\\u037A-\\u037D\\u037F\\u0386\\u0388-\\u038A\\u038C\\u038E-\\u03A1\\u03A3-\\u03F5\\u03F7-\\u0481\\u048A-\\u052F\\u0531-\\u0556\\u0559\\u0560-\\u0588\\u05D0-\\u05EA\\u05EF-\\u05F2\\u0620-\\u064A\\u066E-\\u066F\\u0671-\\u06D3\\u06D5\\u06E5-\\u06E6\\u06EE-\\u06EF\\u06FA-\\u06FC\\u06FF\\u0710\\u0712-\\u072F\\u074D-\\u07A5\\u07B1\\u07CA-\\u07EA\\u07F4-\\u07F5\\u07FA\\u0800-\\u0815\\u081A\\u0824\\u0828\\u0840-\\u0858\\u0860-\\u086A\\u0870-\\u0887\\u0889-\\u088F\\u08A0-\\u08C9\\u0904-\\u0939\\u093D\\u0950\\u0958-\\u0961\\u0971-\\u0980\\u0985-\\u098C\\u098F-\\u0990\\u0993-\\u09A8\\u09AA-\\u09B0\\u09B2\\u09B6-\\u09B9\\u09BD\\u09CE\\u09DC-\\u09DD\\u09DF-\\u09E1\\u09F0-\\u09F1\\u09FC\\u0A05-\\u0A0A\\u0A0F-\\u0A10\\u0A13-\\u0A28\\u0A2A-\\u0A30\\u0A32-\\u0A33\\u0A35-\\u0A36\\u0A38-\\u0A39\\u0A59-\\u0A5C\\u0A5E\\u0A72-\\u0A74\\u0A85-\\u0A8D\\u0A8F-\\u0A91\\u0A93-\\u0AA8\\u0AAA-\\u0AB0\\u0AB2-\\u0AB3\\u0AB5-\\u0AB9\\u0ABD\\u0AD0\\u0AE0-\\u0AE1\\u0AF9\\u0B05-\\u0B0C\\u0B0F-\\u0B10\\u0B13-\\u0B28\\u0B2A-\\u0B30\\u0B32-\\u0B33\\u0B35-\\u0B39\\u0B3D\\u0B5C-\\u0B5D\\u0B5F-\\u0B61\\u0B71\\u0B83\\u0B85-\\u0B8A\\u0B8E-\\u0B90\\u0B92-\\u0B95\\u0B99-\\u0B9A\\u0B9C\\u0B9E-\\u0B9F\\u0BA3-\\u0BA4\\u0BA8-\\u0BAA\\u0BAE-\\u0BB9\\u0BD0\\u0C05-\\u0C0C\\u0C0E-\\u0C10\\u0C12-\\u0C28\\u0C2A-\\u0C39\\u0C3D\\u0C58-\\u0C5A\\u0C5C-\\u0C5D\\u0C60-\\u0C61\\u0C80\\u0C85-\\u0C8C\\u0C8E-\\u0C90\\u0C92-\\u0CA8\\u0CAA-\\u0CB3\\u0CB5-\\u0CB9\\u0CBD\\u0CDC-\\u0CDE\\u0CE0-\\u0CE1\\u0CF1-\\u0CF2\\u0D04-\\u0D0C\\u0D0E-\\u0D10\\u0D12-\\u0D3A\\u0D3D\\u0D4E\\u0D54-\\u0D56\\u0D5F-\\u0D61\\u0D7A-\\u0D7F\\u0D85-\\u0D96\\u0D9A-\\u0DB1\\u0DB3-\\u0DBB\\u0DBD\\u0DC0-\\u0DC6\\u0E01-\\u0E30\\u0E32-\\u0E33\\u0E40-\\u0E46\\u0E81-\\u0E82\\u0E84\\u0E86-\\u0E8A\\u0E8C-\\u0EA3\\u0EA5\\u0EA7-\\u0EB0\\u0EB2-\\u0EB3\\u0EBD\\u0EC0-\\u0EC4\\u0EC6\\u0EDC-\\u0EDF\\u0F00\\u0F40-\\u0F47\\u0F49-\\u0F6C\\u0F88-\\u0F8C\\u1000-\\u102A\\u103F\\u1050-\\u1055\\u105A-\\u105D\\u1061\\u1065-\\u1066\\u106E-\\u1070\\u1075-\\u1081\\u108E\\u10A0-\\u10C5\\u10C7\\u10CD\\u10D0-\\u10FA\\u10FC-\\u1248\\u124A-\\u124D\\u1250-\\u1256\\u1258\\u125A-\\u125D\\u1260-\\u1288\\u128A-\\u128D\\u1290-\\u12B0\\u12B2-\\u12B5\\u12B8-\\u12BE\\u12C0\\u12C2-\\u12C5\\u12C8-\\u12D6\\u12D8-\\u1310\\u1312-\\u1315\\u1318-\\u135A\\u1380-\\u138F\\u13A0-\\u13F5\\u13F8-\\u13FD\\u1401-\\u166C\\u166F-\\u167F\\u1681-\\u169A\\u16A0-\\u16EA\\u16F1-\\u16F8\\u1700-\\u1711\\u171F-\\u1731\\u1740-\\u1751\\u1760-\\u176C\\u176E-\\u1770\\u1780-\\u17B3\\u17D7\\u17DC\\u1820-\\u1878\\u1880-\\u1884\\u1887-\\u18A8\\u18AA\\u18B0-\\u18F5\\u1900-\\u191E\\u1950-\\u196D\\u1970-\\u1974\\u1980-\\u19AB\\u19B0-\\u19C9\\u1A00-\\u1A16\\u1A20-\\u1A54\\u1AA7\\u1B05-\\u1B33\\u1B45-\\u1B4C\\u1B83-\\u1BA0\\u1BAE-\\u1BAF\\u1BBA-\\u1BE5\\u1C00-\\u1C23\\u1C4D-\\u1C4F\\u1C5A-\\u1C7D\\u1C80-\\u1C8A\\u1C90-\\u1CBA\\u1CBD-\\u1CBF\\u1CE9-\\u1CEC\\u1CEE-\\u1CF3\\u1CF5-\\u1CF6\\u1CFA\\u1D00-\\u1DBF\\u1E00-\\u1F15\\u1F18-\\u1F1D\\u1F20-\\u1F45\\u1F48-\\u1F4D\\u1F50-\\u1F57\\u1F59\\u1F5B\\u1F5D\\u1F5F-\\u1F7D\\u1F80-\\u1FB4\\u1FB6-\\u1FBC\\u1FBE\\u1FC2-\\u1FC4\\u1FC6-\\u1FCC\\u1FD0-\\u1FD3\\u1FD6-\\u1FDB\\u1FE0-\\u1FEC\\u1FF2-\\u1FF4\\u1FF6-\\u1FFC\\u2071\\u207F\\u2090-\\u209C\\u2102\\u2107\\u210A-\\u2113\\u2115\\u2119-\\u211D\\u2124\\u2126\\u2128\\u212A-\\u212D\\u212F-\\u2139\\u213C-\\u213F\\u2145-\\u2149\\u214E\\u2183-\\u2184\\u2C00-\\u2CE4\\u2CEB-\\u2CEE\\u2CF2-\\u2CF3\\u2D00-\\u2D25\\u2D27\\u2D2D\\u2D30-\\u2D67\\u2D6F\\u2D80-\\u2D96\\u2DA0-\\u2DA6\\u2DA8-\\u2DAE\\u2DB0-\\u2DB6\\u2DB8-\\u2DBE\\u2DC0-\\u2DC6\\u2DC8-\\u2DCE\\u2DD0-\\u2DD6\\u2DD8-\\u2DDE\\u2E2F\\u3005-\\u3006\\u3031-\\u3035\\u303B-\\u303C\\u3041-\\u3096\\u309D-\\u309F\\u30A1-\\u30FA\\u30FC-\\u30FF\\u3105-\\u312F\\u3131-\\u318E\\u31A0-\\u31BF\\u31F0-\\u31FF\\u3400-\\u4DBF\\u4E00-\\uA48C\\uA4D0-\\uA4FD\\uA500-\\uA60C\\uA610-\\uA61F\\uA62A-\\uA62B\\uA640-\\uA66E\\uA67F-\\uA69D\\uA6A0-\\uA6E5\\uA717-\\uA71F\\uA722-\\uA788\\uA78B-\\uA7DC\\uA7F1-\\uA801\\uA803-\\uA805\\uA807-\\uA80A\\uA80C-\\uA822\\uA840-\\uA873\\uA882-\\uA8B3\\uA8F2-\\uA8F7\\uA8FB\\uA8FD-\\uA8FE\\uA90A-\\uA925\\uA930-\\uA946\\uA960-\\uA97C\\uA984-\\uA9B2\\uA9CF\\uA9E0-\\uA9E4\\uA9E6-\\uA9EF\\uA9FA-\\uA9FE\\uAA00-\\uAA28\\uAA40-\\uAA42\\uAA44-\\uAA4B\\uAA60-\\uAA76\\uAA7A\\uAA7E-\\uAAAF\\uAAB1\\uAAB5-\\uAAB6\\uAAB9-\\uAABD\\uAAC0\\uAAC2\\uAADB-\\uAADD\\uAAE0-\\uAAEA\\uAAF2-\\uAAF4\\uAB01-\\uAB06\\uAB09-\\uAB0E\\uAB11-\\uAB16\\uAB20-\\uAB26\\uAB28-\\uAB2E\\uAB30-\\uAB5A\\uAB5C-\\uAB69\\uAB70-\\uABE2\\uAC00-\\uD7A3\\uD7B0-\\uD7C6\\uD7CB-\\uD7FB\\uF900-\\uFA6D\\uFA70-\\uFAD9\\uFB00-\\uFB06\\uFB13-\\uFB17\\uFB1D\\uFB1F-\\uFB28\\uFB2A-\\uFB36\\uFB38-\\uFB3C\\uFB3E\\uFB40-\\uFB41\\uFB43-\\uFB44\\uFB46-\\uFBB1\\uFBD3-\\uFD3D\\uFD50-\\uFD8F\\uFD92-\\uFDC7\\uFDF0-\\uFDFB\\uFE70-\\uFE74\\uFE76-\\uFEFC\\uFF21-\\uFF3A\\uFF41-\\uFF5A\\uFF66-\\uFFBE\\uFFC2-\\uFFC7\\uFFCA-\\uFFCF\\uFFD2-\\uFFD7\\uFFDA-\\uFFDC\\u{10000}-\\u{1000B}\\u{1000D}-\\u{10026}\\u{10028}-\\u{1003A}\\u{1003C}-\\u{1003D}\\u{1003F}-\\u{1004D}\\u{10050}-\\u{1005D}\\u{10080}-\\u{100FA}\\u{10280}-\\u{1029C}\\u{102A0}-\\u{102D0}\\u{10300}-\\u{1031F}\\u{1032D}-\\u{10340}\\u{10342}-\\u{10349}\\u{10350}-\\u{10375}\\u{10380}-\\u{1039D}\\u{103A0}-\\u{103C3}\\u{103C8}-\\u{103CF}\\u{10400}-\\u{1049D}\\u{104B0}-\\u{104D3}\\u{104D8}-\\u{104FB}\\u{10500}-\\u{10527}\\u{10530}-\\u{10563}\\u{10570}-\\u{1057A}\\u{1057C}-\\u{1058A}\\u{1058C}-\\u{10592}\\u{10594}-\\u{10595}\\u{10597}-\\u{105A1}\\u{105A3}-\\u{105B1}\\u{105B3}-\\u{105B9}\\u{105BB}-\\u{105BC}\\u{105C0}-\\u{105F3}\\u{10600}-\\u{10736}\\u{10740}-\\u{10755}\\u{10760}-\\u{10767}\\u{10780}-\\u{10785}\\u{10787}-\\u{107B0}\\u{107B2}-\\u{107BA}\\u{10800}-\\u{10805}\\u{10808}\\u{1080A}-\\u{10835}\\u{10837}-\\u{10838}\\u{1083C}\\u{1083F}-\\u{10855}\\u{10860}-\\u{10876}\\u{10880}-\\u{1089E}\\u{108E0}-\\u{108F2}\\u{108F4}-\\u{108F5}\\u{10900}-\\u{10915}\\u{10920}-\\u{10939}\\u{10940}-\\u{10959}\\u{10980}-\\u{109B7}\\u{109BE}-\\u{109BF}\\u{10A00}\\u{10A10}-\\u{10A13}\\u{10A15}-\\u{10A17}\\u{10A19}-\\u{10A35}\\u{10A60}-\\u{10A7C}\\u{10A80}-\\u{10A9C}\\u{10AC0}-\\u{10AC7}\\u{10AC9}-\\u{10AE4}\\u{10B00}-\\u{10B35}\\u{10B40}-\\u{10B55}\\u{10B60}-\\u{10B72}\\u{10B80}-\\u{10B91}\\u{10C00}-\\u{10C48}\\u{10C80}-\\u{10CB2}\\u{10CC0}-\\u{10CF2}\\u{10D00}-\\u{10D23}\\u{10D4A}-\\u{10D65}\\u{10D6F}-\\u{10D85}\\u{10E80}-\\u{10EA9}\\u{10EB0}-\\u{10EB1}\\u{10EC2}-\\u{10EC7}\\u{10F00}-\\u{10F1C}\\u{10F27}\\u{10F30}-\\u{10F45}\\u{10F70}-\\u{10F81}\\u{10FB0}-\\u{10FC4}\\u{10FE0}-\\u{10FF6}\\u{11003}-\\u{11037}\\u{11071}-\\u{11072}\\u{11075}\\u{11083}-\\u{110AF}\\u{110D0}-\\u{110E8}\\u{11103}-\\u{11126}\\u{11144}\\u{11147}\\u{11150}-\\u{11172}\\u{11176}\\u{11183}-\\u{111B2}\\u{111C1}-\\u{111C4}\\u{111DA}\\u{111DC}\\u{11200}-\\u{11211}\\u{11213}-\\u{1122B}\\u{1123F}-\\u{11240}\\u{11280}-\\u{11286}\\u{11288}\\u{1128A}-\\u{1128D}\\u{1128F}-\\u{1129D}\\u{1129F}-\\u{112A8}\\u{112B0}-\\u{112DE}\\u{11305}-\\u{1130C}\\u{1130F}-\\u{11310}\\u{11313}-\\u{11328}\\u{1132A}-\\u{11330}\\u{11332}-\\u{11333}\\u{11335}-\\u{11339}\\u{1133D}\\u{11350}\\u{1135D}-\\u{11361}\\u{11380}-\\u{11389}\\u{1138B}\\u{1138E}\\u{11390}-\\u{113B5}\\u{113B7}\\u{113D1}\\u{113D3}\\u{11400}-\\u{11434}\\u{11447}-\\u{1144A}\\u{1145F}-\\u{11461}\\u{11480}-\\u{114AF}\\u{114C4}-\\u{114C5}\\u{114C7}\\u{11580}-\\u{115AE}\\u{115D8}-\\u{115DB}\\u{11600}-\\u{1162F}\\u{11644}\\u{11680}-\\u{116AA}\\u{116B8}\\u{11700}-\\u{1171A}\\u{11740}-\\u{11746}\\u{11800}-\\u{1182B}\\u{118A0}-\\u{118DF}\\u{118FF}-\\u{11906}\\u{11909}\\u{1190C}-\\u{11913}\\u{11915}-\\u{11916}\\u{11918}-\\u{1192F}\\u{1193F}\\u{11941}\\u{119A0}-\\u{119A7}\\u{119AA}-\\u{119D0}\\u{119E1}\\u{119E3}\\u{11A00}\\u{11A0B}-\\u{11A32}\\u{11A3A}\\u{11A50}\\u{11A5C}-\\u{11A89}\\u{11A9D}\\u{11AB0}-\\u{11AF8}\\u{11BC0}-\\u{11BE0}\\u{11C00}-\\u{11C08}\\u{11C0A}-\\u{11C2E}\\u{11C40}\\u{11C72}-\\u{11C8F}\\u{11D00}-\\u{11D06}\\u{11D08}-\\u{11D09}\\u{11D0B}-\\u{11D30}\\u{11D46}\\u{11D60}-\\u{11D65}\\u{11D67}-\\u{11D68}\\u{11D6A}-\\u{11D89}\\u{11D98}\\u{11DB0}-\\u{11DDB}\\u{11EE0}-\\u{11EF2}\\u{11F02}\\u{11F04}-\\u{11F10}\\u{11F12}-\\u{11F33}\\u{11FB0}\\u{12000}-\\u{12399}\\u{12480}-\\u{12543}\\u{12F90}-\\u{12FF0}\\u{13000}-\\u{1342F}\\u{13441}-\\u{13446}\\u{13460}-\\u{143FA}\\u{14400}-\\u{14646}\\u{16100}-\\u{1611D}\\u{16800}-\\u{16A38}\\u{16A40}-\\u{16A5E}\\u{16A70}-\\u{16ABE}\\u{16AD0}-\\u{16AED}\\u{16B00}-\\u{16B2F}\\u{16B40}-\\u{16B43}\\u{16B63}-\\u{16B77}\\u{16B7D}-\\u{16B8F}\\u{16D40}-\\u{16D6C}\\u{16E40}-\\u{16E7F}\\u{16EA0}-\\u{16EB8}\\u{16EBB}-\\u{16ED3}\\u{16F00}-\\u{16F4A}\\u{16F50}\\u{16F93}-\\u{16F9F}\\u{16FE0}-\\u{16FE1}\\u{16FE3}\\u{16FF2}-\\u{16FF3}\\u{17000}-\\u{18CD5}\\u{18CFF}-\\u{18D1E}\\u{18D80}-\\u{18DF2}\\u{1AFF0}-\\u{1AFF3}\\u{1AFF5}-\\u{1AFFB}\\u{1AFFD}-\\u{1AFFE}\\u{1B000}-\\u{1B122}\\u{1B132}\\u{1B150}-\\u{1B152}\\u{1B155}\\u{1B164}-\\u{1B167}\\u{1B170}-\\u{1B2FB}\\u{1BC00}-\\u{1BC6A}\\u{1BC70}-\\u{1BC7C}\\u{1BC80}-\\u{1BC88}\\u{1BC90}-\\u{1BC99}\\u{1D400}-\\u{1D454}\\u{1D456}-\\u{1D49C}\\u{1D49E}-\\u{1D49F}\\u{1D4A2}\\u{1D4A5}-\\u{1D4A6}\\u{1D4A9}-\\u{1D4AC}\\u{1D4AE}-\\u{1D4B9}\\u{1D4BB}\\u{1D4BD}-\\u{1D4C3}\\u{1D4C5}-\\u{1D505}\\u{1D507}-\\u{1D50A}\\u{1D50D}-\\u{1D514}\\u{1D516}-\\u{1D51C}\\u{1D51E}-\\u{1D539}\\u{1D53B}-\\u{1D53E}\\u{1D540}-\\u{1D544}\\u{1D546}\\u{1D54A}-\\u{1D550}\\u{1D552}-\\u{1D6A5}\\u{1D6A8}-\\u{1D6C0}\\u{1D6C2}-\\u{1D6DA}\\u{1D6DC}-\\u{1D6FA}\\u{1D6FC}-\\u{1D714}\\u{1D716}-\\u{1D734}\\u{1D736}-\\u{1D74E}\\u{1D750}-\\u{1D76E}\\u{1D770}-\\u{1D788}\\u{1D78A}-\\u{1D7A8}\\u{1D7AA}-\\u{1D7C2}\\u{1D7C4}-\\u{1D7CB}\\u{1DF00}-\\u{1DF1E}\\u{1DF25}-\\u{1DF2A}\\u{1E030}-\\u{1E06D}\\u{1E100}-\\u{1E12C}\\u{1E137}-\\u{1E13D}\\u{1E14E}\\u{1E290}-\\u{1E2AD}\\u{1E2C0}-\\u{1E2EB}\\u{1E4D0}-\\u{1E4EB}\\u{1E5D0}-\\u{1E5ED}\\u{1E5F0}\\u{1E6C0}-\\u{1E6DE}\\u{1E6E0}-\\u{1E6E2}\\u{1E6E4}-\\u{1E6E5}\\u{1E6E7}-\\u{1E6ED}\\u{1E6F0}-\\u{1E6F4}\\u{1E6FE}-\\u{1E6FF}\\u{1E7E0}-\\u{1E7E6}\\u{1E7E8}-\\u{1E7EB}\\u{1E7ED}-\\u{1E7EE}\\u{1E7F0}-\\u{1E7FE}\\u{1E800}-\\u{1E8C4}\\u{1E900}-\\u{1E943}\\u{1E94B}\\u{1EE00}-\\u{1EE03}\\u{1EE05}-\\u{1EE1F}\\u{1EE21}-\\u{1EE22}\\u{1EE24}\\u{1EE27}\\u{1EE29}-\\u{1EE32}\\u{1EE34}-\\u{1EE37}\\u{1EE39}\\u{1EE3B}\\u{1EE42}\\u{1EE47}\\u{1EE49}\\u{1EE4B}\\u{1EE4D}-\\u{1EE4F}\\u{1EE51}-\\u{1EE52}\\u{1EE54}\\u{1EE57}\\u{1EE59}\\u{1EE5B}\\u{1EE5D}\\u{1EE5F}\\u{1EE61}-\\u{1EE62}\\u{1EE64}\\u{1EE67}-\\u{1EE6A}\\u{1EE6C}-\\u{1EE72}\\u{1EE74}-\\u{1EE77}\\u{1EE79}-\\u{1EE7C}\\u{1EE7E}\\u{1EE80}-\\u{1EE89}\\u{1EE8B}-\\u{1EE9B}\\u{1EEA1}-\\u{1EEA3}\\u{1EEA5}-\\u{1EEA9}\\u{1EEAB}-\\u{1EEBB}\\u{20000}-\\u{2A6DF}\\u{2A700}-\\u{2B81D}\\u{2B820}-\\u{2CEAD}\\u{2CEB0}-\\u{2EBE0}\\u{2EBF0}-\\u{2EE5D}\\u{2F800}-\\u{2FA1D}\\u{30000}-\\u{3134A}\\u{31350}-\\u{33479}\\u0030-\\u0039\\u00B2-\\u00B3\\u00B9\\u00BC-\\u00BE\\u0660-\\u0669\\u06F0-\\u06F9\\u07C0-\\u07C9\\u0966-\\u096F\\u09E6-\\u09EF\\u09F4-\\u09F9\\u0A66-\\u0A6F\\u0AE6-\\u0AEF\\u0B66-\\u0B6F\\u0B72-\\u0B77\\u0BE6-\\u0BF2\\u0C66-\\u0C6F\\u0C78-\\u0C7E\\u0CE6-\\u0CEF\\u0D58-\\u0D5E\\u0D66-\\u0D78\\u0DE6-\\u0DEF\\u0E50-\\u0E59\\u0ED0-\\u0ED9\\u0F20-\\u0F33\\u1040-\\u1049\\u1090-\\u1099\\u1369-\\u137C\\u16EE-\\u16F0\\u17E0-\\u17E9\\u17F0-\\u17F9\\u1810-\\u1819\\u1946-\\u194F\\u19D0-\\u19DA\\u1A80-\\u1A89\\u1A90-\\u1A99\\u1B50-\\u1B59\\u1BB0-\\u1BB9\\u1C40-\\u1C49\\u1C50-\\u1C59\\u2070\\u2074-\\u2079\\u2080-\\u2089\\u2150-\\u2182\\u2185-\\u2189\\u2460-\\u249B\\u24EA-\\u24FF\\u2776-\\u2793\\u2CFD\\u3007\\u3021-\\u3029\\u3038-\\u303A\\u3192-\\u3195\\u3220-\\u3229\\u3248-\\u324F\\u3251-\\u325F\\u3280-\\u3289\\u32B1-\\u32BF\\uA620-\\uA629\\uA6E6-\\uA6EF\\uA830-\\uA835\\uA8D0-\\uA8D9\\uA900-\\uA909\\uA9D0-\\uA9D9\\uA9F0-\\uA9F9\\uAA50-\\uAA59\\uABF0-\\uABF9\\uFF10-\\uFF19\\u{10107}-\\u{10133}\\u{10140}-\\u{10178}\\u{1018A}-\\u{1018B}\\u{102E1}-\\u{102FB}\\u{10320}-\\u{10323}\\u{10341}\\u{1034A}\\u{103D1}-\\u{103D5}\\u{104A0}-\\u{104A9}\\u{10858}-\\u{1085F}\\u{10879}-\\u{1087F}\\u{108A7}-\\u{108AF}\\u{108FB}-\\u{108FF}\\u{10916}-\\u{1091B}\\u{109BC}-\\u{109BD}\\u{109C0}-\\u{109CF}\\u{109D2}-\\u{109FF}\\u{10A40}-\\u{10A48}\\u{10A7D}-\\u{10A7E}\\u{10A9D}-\\u{10A9F}\\u{10AEB}-\\u{10AEF}\\u{10B58}-\\u{10B5F}\\u{10B78}-\\u{10B7F}\\u{10BA9}-\\u{10BAF}\\u{10CFA}-\\u{10CFF}\\u{10D30}-\\u{10D39}\\u{10D40}-\\u{10D49}\\u{10E60}-\\u{10E7E}\\u{10F1D}-\\u{10F26}\\u{10F51}-\\u{10F54}\\u{10FC5}-\\u{10FCB}\\u{11052}-\\u{1106F}\\u{110F0}-\\u{110F9}\\u{11136}-\\u{1113F}\\u{111D0}-\\u{111D9}\\u{111E1}-\\u{111F4}\\u{112F0}-\\u{112F9}\\u{11450}-\\u{11459}\\u{114D0}-\\u{114D9}\\u{11650}-\\u{11659}\\u{116C0}-\\u{116C9}\\u{116D0}-\\u{116E3}\\u{11730}-\\u{1173B}\\u{118E0}-\\u{118F2}\\u{11950}-\\u{11959}\\u{11BF0}-\\u{11BF9}\\u{11C50}-\\u{11C6C}\\u{11D50}-\\u{11D59}\\u{11DA0}-\\u{11DA9}\\u{11DE0}-\\u{11DE9}\\u{11F50}-\\u{11F59}\\u{11FC0}-\\u{11FD4}\\u{12400}-\\u{1246E}\\u{16130}-\\u{16139}\\u{16A60}-\\u{16A69}\\u{16AC0}-\\u{16AC9}\\u{16B50}-\\u{16B59}\\u{16B5B}-\\u{16B61}\\u{16D70}-\\u{16D79}\\u{16E80}-\\u{16E96}\\u{16FF4}-\\u{16FF6}\\u{1CCF0}-\\u{1CCF9}\\u{1D2C0}-\\u{1D2D3}\\u{1D2E0}-\\u{1D2F3}\\u{1D360}-\\u{1D378}\\u{1D7CE}-\\u{1D7FF}\\u{1E140}-\\u{1E149}\\u{1E2F0}-\\u{1E2F9}\\u{1E4F0}-\\u{1E4F9}\\u{1E5F1}-\\u{1E5FA}\\u{1E8C7}-\\u{1E8CF}\\u{1E950}-\\u{1E959}\\u{1EC71}-\\u{1ECAB}\\u{1ECAD}-\\u{1ECAF}\\u{1ECB1}-\\u{1ECB4}\\u{1ED01}-\\u{1ED2D}\\u{1ED2F}-\\u{1ED3D}\\u{1F100}-\\u{1F10C}\\u{1FBF0}-\\u{1FBF9}]+", "gu"))) != null ? _text_match : []).length;
+};
+/** One verdict from what became of the home page and how much of the site was turned away. */ function readVerdictOf(home, blocked, total) {
+    var _ref, _ref1;
+    const edge = (home == null ? void 0 : home.edge) === 'vercel' ? 'vercel' : (home == null ? void 0 : home.edge) === 'cloudflare' ? 'cloudflare' : null;
+    const base = {
+        blocked,
+        total,
+        homeStatus: (_ref = home == null ? void 0 : home.status) != null ? _ref : null,
+        homeProblem: (_ref1 = home == null ? void 0 : home.problem) != null ? _ref1 : null,
+        edge
+    };
+    if (home) return _extends({
+        kind: home.problem === 'blocked' ? 'home-blocked' : 'home-unreadable'
+    }, base);
+    return _extends({
+        kind: total > 0 && blocked * 2 > total ? 'mostly-blocked' : 'ok'
+    }, base);
+}
+/** What a crawler skip of the home page means for it. */ function homeProblemOfSkip(reason, status) {
+    switch(reason){
+        case 'http-error':
+            return status !== undefined && AI_SITE_IMPORT_BLOCKED_STATUSES.has(status) ? 'blocked' : 'http-error';
+        case 'off-origin-redirect':
+            return 'off-site';
+        case 'not-html':
+        case 'too-large':
+        case 'robots':
+        case 'login-form':
+        case 'not-reached':
+            return reason;
+        default:
+            return 'unreachable';
+    }
+}
+/**
+ * The read's verdict (AGL-4083), pure over the crawl: whether the home page
+ * came in, and how much of the site bot protection turned away. An import
+ * whose verdict stops (`aiSiteImportReadStops`) plans no wave, quotes
+ * nothing and spends nothing: the person is told why and checks again.
+ */ export function aiSiteImportReadVerdict(crawled) {
+    var _crawled_pages_find, _skip_status, _skip_edge;
+    const blocked = crawled.skipped.filter((entry)=>entry.reason === 'http-error' && entry.status !== undefined && AI_SITE_IMPORT_BLOCKED_STATUSES.has(entry.status)).length;
+    const total = Math.max(crawled.sitemapUrls, crawled.pages.length + crawled.skipped.length);
+    const page = (_crawled_pages_find = crawled.pages.find((entry)=>entry.foundBy === 'home')) != null ? _crawled_pages_find : crawled.pages.find((entry)=>(entry.path.replace(/\/+$/, '') || '/') === '/');
+    if (page) {
+        const words = page.blocks.reduce((sum, block)=>{
+            var _block_text;
+            return sum + (block.kind === 'image' || block.inline ? 0 : wordsOf((_block_text = block.text) != null ? _block_text : ''));
+        }, 0);
+        const empty = words < AI_SITE_IMPORT_MIN_WORDS && page.images.length === 0;
+        return readVerdictOf(empty ? {
+            problem: 'empty',
+            status: null,
+            edge: null
+        } : null, blocked, total);
+    }
+    const skip = crawled.skipped.find((entry)=>pathOfUrl(entry.url) === '/');
+    return readVerdictOf(skip ? {
+        problem: homeProblemOfSkip(skip.reason, skip.status),
+        status: (_skip_status = skip.status) != null ? _skip_status : null,
+        edge: (_skip_edge = skip.edge) != null ? _skip_edge : null
+    } : {
+        problem: 'not-reached',
+        status: null,
+        edge: null
+    }, blocked, total);
+}
+/** Whether a verdict stops the import before anything is planned: its home page is not in. */ export function aiSiteImportReadStops(verdict) {
+    return (verdict == null ? void 0 : verdict.kind) === 'home-blocked' || (verdict == null ? void 0 : verdict.kind) === 'home-unreadable';
+}
+/** The review an import that stopped at its read waits under: `blocked` or `unreadable`; `null` when it goes on. */ export function aiSiteImportReadVerdictReason(verdict) {
+    if ((verdict == null ? void 0 : verdict.kind) === 'home-blocked') return 'blocked';
+    if ((verdict == null ? void 0 : verdict.kind) === 'home-unreadable') return 'unreadable';
+    return null;
+}
+/** Where a site owner lets AglynBot through, on the edges that name themselves. */ export const AI_SITE_IMPORT_EDGE_HINTS = {
+    vercel: 'Your site answers through Vercel: in its Firewall, under Rules, add a bypass rule for the user agent AglynBot.',
+    cloudflare: 'Your site answers through Cloudflare: under Security, WAF, add a custom rule that skips the checks when the user agent contains AglynBot.'
+};
+const NOTHING_SPENT = 'Nothing was brought in and no AI credits were used.';
+/** Why the home page could not be read, inside the sentence that says so. */ function homeProblemText(verdict) {
+    switch(verdict.homeProblem){
+        case 'http-error':
+            return verdict.homeStatus ? `it answered with an error, ${verdict.homeStatus}` : 'it answered with an error';
+        case 'not-html':
+            return 'it did not answer with a web page';
+        case 'too-large':
+            return 'it is larger than a page can be to bring in';
+        case 'off-site':
+            return 'it sends visitors to another website';
+        case 'not-reached':
+            return 'the read ran out of time before it';
+        default:
+            return 'it did not answer';
+    }
+}
+/**
+ * What the person is told about the read (AGL-4083), in console words: why
+ * nothing was brought in and what to do, or that most of the site was turned
+ * away. `null` when there is nothing to say.
+ */ export function aiSiteImportReadVerdictMessage(verdict, domain) {
+    if (!verdict) return null;
+    if (verdict.kind === 'home-blocked') {
+        const others = Math.max(0, verdict.blocked - 1);
+        const more = others ? `, and ${others.toLocaleString('en-US')} other ${others === 1 ? 'page' : 'pages'}` : '';
+        const lead = `Your site’s bot protection or firewall turned AglynBot away before it could read your home page${more}. ${NOTHING_SPENT} Allow AglynBot in your site’s bot or firewall settings, then choose Check again.`;
+        return verdict.edge ? `${lead} ${AI_SITE_IMPORT_EDGE_HINTS[verdict.edge]}` : lead;
+    }
+    if (verdict.kind === 'home-unreadable') {
+        switch(verdict.homeProblem){
+            case 'robots':
+                return `Your site’s robots.txt asks AglynBot not to read your home page. ${NOTHING_SPENT} Allow AglynBot in your robots.txt, then choose Check again.`;
+            case 'login-form':
+                return `Your home page asks for a password, and pages that do are never brought in. ${NOTHING_SPENT} Open your home page to visitors, then choose Check again.`;
+            case 'empty':
+                return `No text could be read from your home page. If your site is built in a way that needs a browser to show its words, it cannot be brought in yet. ${NOTHING_SPENT}`;
+            default:
+                return `Your home page could not be read (${homeProblemText(verdict)}). ${NOTHING_SPENT} Check that ${domain} is up and answers at https://${domain}/, then choose Check again.`;
+        }
+    }
+    if (verdict.kind === 'mostly-blocked') {
+        return `${verdict.blocked.toLocaleString('en-US')} of ${verdict.total.toLocaleString('en-US')} pages were turned away by your site’s bot protection; see Not brought in.`;
+    }
+    return null;
+}
+/** The verdict an inventory reads from its own pages and skips, when the step did not hand it the crawl's. */ function inventoryVerdict(input) {
+    var _input_skippedOverflow, _input_pages_find;
+    const blocked = input.skipped.filter((entry)=>entry.reason === 'blocked').length;
+    const total = input.pages.length + input.skipped.length + Math.max(0, (_input_skippedOverflow = input.skippedOverflow) != null ? _input_skippedOverflow : 0);
+    const page = (_input_pages_find = input.pages.find((entry)=>entry.foundBy === 'home')) != null ? _input_pages_find : input.pages.find((entry)=>(entry.path.replace(/\/+$/, '') || '/') === '/');
+    if (page) {
+        const empty = page.words < AI_SITE_IMPORT_MIN_WORDS && page.images === 0;
+        return readVerdictOf(empty ? {
+            problem: 'empty',
+            status: null,
+            edge: null
+        } : null, blocked, total);
+    }
+    const skip = input.skipped.find((entry)=>pathOfUrl(entry.url) === '/');
+    const problem = !skip ? 'not-reached' : skip.reason === 'blocked' ? 'blocked' : skip.reason === 'robots' || skip.reason === 'login-form' || skip.reason === 'too-large' || skip.reason === 'off-site' || skip.reason === 'not-reached' || skip.reason === 'empty' ? skip.reason : 'unreachable';
+    return readVerdictOf({
+        problem,
+        status: null,
+        edge: null
+    }, blocked, total);
+}
 /*------------------------------------------
  * Grouping and waves
  *-----------------------------------------*/ const clipLabel = (text)=>{
@@ -203,9 +395,10 @@ const segmentsOf = (path)=>String(path != null ? path : '').split('/').filter(Bo
  * The site's inventory cut into waves, every page left out said with its
  * reason. Pure: the step reads the site, this decides what comes in.
  */ export function aiSiteImportInventory(input) {
-    var _input_wavePages, _leftOutCount_unreadable, _input_reviewId;
+    var _input_wavePages, _leftOutCount_unreadable, _input_verdict, _input_reviewId;
     var _waves_;
     const wavePages = Math.max(1, Math.floor((_input_wavePages = input.wavePages) != null ? _input_wavePages : AI_SITE_IMPORT_WAVE_PAGES));
+    const limit = input.pageLimit === null || !Number.isFinite(input.pageLimit) ? null : Math.max(0, Math.floor(input.pageLimit));
     const leftOut = [];
     const leftOutCount = {};
     const leave = (entry)=>{
@@ -213,6 +406,9 @@ const segmentsOf = (path)=>String(path != null ? path : '').split('/').filter(Bo
         leftOutCount[entry.reason] = ((_leftOutCount_entry_reason = leftOutCount[entry.reason]) != null ? _leftOutCount_entry_reason : 0) + 1;
         if (leftOut.length < AI_SITE_IMPORT_LEFT_OUT_LISTED) leftOut.push(entry);
     };
+    // The read asks for a wave's worth past the plan's room, so a page its page
+    // budget never reached is past the plan; any other stop is not the plan's.
+    const unreached = input.stoppedBy === 'page-budget' && limit !== null ? 'plan-limit' : 'not-reached';
     for (const skipped of input.skipped){
         let path = '/';
         try {
@@ -224,10 +420,33 @@ const segmentsOf = (path)=>String(path != null ? path : '').split('/').filter(Bo
             url: skipped.url,
             path,
             title: '',
-            reason: skipped.reason
+            reason: skipped.reason === 'not-reached' ? unreached : skipped.reason
         });
     }
     if (input.skippedOverflow) leftOutCount['unreadable'] = ((_leftOutCount_unreadable = leftOutCount['unreadable']) != null ? _leftOutCount_unreadable : 0) + input.skippedOverflow;
+    // Home first: a site read without its home page plans nothing at all.
+    const verdict = (_input_verdict = input.verdict) != null ? _input_verdict : inventoryVerdict(input);
+    if (aiSiteImportReadStops(verdict)) {
+        var _input_reviewId1;
+        return {
+            domain: input.domain,
+            siteName: clipLabel(input.siteName) || input.domain,
+            ownership: input.ownership,
+            readAt: input.readAt,
+            pageLimit: limit,
+            existingPages: Math.max(0, input.existingPages),
+            found: input.pages.length,
+            stoppedBy: input.stoppedBy,
+            groups: [],
+            pages: [],
+            waves: [],
+            leftOut,
+            leftOutCount,
+            frame: null,
+            reviewId: (_input_reviewId1 = input.reviewId) != null ? _input_reviewId1 : null,
+            verdict
+        };
+    }
     const readable = input.pages.filter((page)=>{
         if (page.words >= AI_SITE_IMPORT_MIN_WORDS || page.images > 0) return true;
         leave({
@@ -239,7 +458,6 @@ const segmentsOf = (path)=>String(path != null ? path : '').split('/').filter(Bo
         return false;
     });
     const { groups, groupOf } = aiSiteImportGroups(readable, input.menu);
-    const limit = input.pageLimit === null || !Number.isFinite(input.pageLimit) ? null : Math.max(0, Math.floor(input.pageLimit));
     let room = limit === null ? Number.POSITIVE_INFINITY : Math.max(0, limit - Math.max(0, input.existingPages));
     const pages = [];
     const waves = [];
@@ -311,7 +529,8 @@ const segmentsOf = (path)=>String(path != null ? path : '').split('/').filter(Bo
         frame: ((_waves_ = waves[0]) == null ? void 0 : _waves_.frame) ? {
             draftId: input.mintId()
         } : null,
-        reviewId: (_input_reviewId = input.reviewId) != null ? _input_reviewId : null
+        reviewId: (_input_reviewId = input.reviewId) != null ? _input_reviewId : null,
+        verdict
     };
 }
 /** The item ledger an inventory starts: the header and footer, then every page, all pending. */ export function aiSiteImportLedger(state) {
@@ -346,7 +565,7 @@ const OPEN = new Set([
 }
 /** The next wave a person may start, else `null`. */ export function aiSiteImportNextWave(state) {
     var _state_waves_find;
-    if (!state || aiSiteImportRunningWave(state)) return null;
+    if (!state || aiSiteImportRunningWave(state) || aiSiteImportReadStops(state.verdict)) return null;
     return (_state_waves_find = state.waves.find((wave)=>wave.status === 'waiting')) != null ? _state_waves_find : null;
 }
 /** The wave a person is reviewing, else `null`. */ export function aiSiteImportReviewWave(state) {
@@ -354,7 +573,8 @@ const OPEN = new Set([
     return (_ref = state == null ? void 0 : state.waves.find((wave)=>wave.status === 'review')) != null ? _ref : null;
 }
 /** What the next pass of the step does. */ export function aiSiteImportNextWork(state, items) {
-    if (!state) return {
+    // A read that stopped at its home page reads again: the person's Check again.
+    if (!state || aiSiteImportReadStops(state.verdict)) return {
         kind: 'inventory'
     };
     const running = aiSiteImportRunningWave(state);
@@ -420,15 +640,62 @@ const withWave = (state, index, patch)=>_extends({}, state, {
     });
     return aiSiteImportPagesRead(next, reads);
 }
-/** More of a running wave's pages, read now. */ export function aiSiteImportPagesRead(state, reads) {
+/** How many characters a block holds, as the job document stores it. */ export function aiSiteImportBlockChars(block) {
+    if (block.kind === 'image') return block.url.length + block.alt.length;
+    return block.text.length + (block.kind === 'link' ? block.href.length : 0);
+}
+/** A page's blocks in order, as many as fit `maxChars` and `maxBlocks`, and whether any had to go. */ export function aiSiteImportCapBlocks(blocks, maxChars, maxBlocks = AI_SITE_IMPORT_READ_MAX_BLOCKS) {
+    const kept = [];
+    let chars = 0;
+    for (const block of blocks){
+        const size = aiSiteImportBlockChars(block);
+        if (kept.length >= maxBlocks || chars + size > maxChars) return {
+            blocks: kept,
+            chars,
+            cut: true
+        };
+        kept.push(block);
+        chars += size;
+    }
+    return {
+        blocks: kept,
+        chars,
+        cut: false
+    };
+}
+const readChars = (read)=>{
+    var _ref;
+    return ((_ref = read == null ? void 0 : read.blocks) != null ? _ref : []).reduce((total, block)=>total + aiSiteImportBlockChars(block), 0);
+};
+/**
+ * More of a running wave's pages, read now. Their blocks share
+ * `AI_SITE_IMPORT_WAVE_READ_MAX_CHARS` with every read still held: a page
+ * past it keeps what fits, in order, and is marked `cut`.
+ */ export function aiSiteImportPagesRead(state, reads) {
+    let held = state.pages.filter((page)=>!reads.has(page.slot)).reduce((total, page)=>total + readChars(page.read), 0);
     let next = state;
     for (const [slot, read] of reads){
         var _next_pages_find;
-        next = typeof read === 'string' ? withPage(next, slot, {
-            refused: read,
-            read: null
-        }) : withPage(next, slot, {
-            read,
+        if (typeof read === 'string') {
+            next = withPage(next, slot, {
+                refused: read,
+                read: null
+            });
+            continue;
+        }
+        let kept = read;
+        if (read.blocks) {
+            const room = Math.max(0, Math.min(AI_SITE_IMPORT_READ_MAX_CHARS, AI_SITE_IMPORT_WAVE_READ_MAX_CHARS - held));
+            const capped = aiSiteImportCapBlocks(read.blocks, room);
+            held += capped.chars;
+            kept = _extends({}, read, {
+                blocks: capped.blocks
+            }, capped.cut || read.cut ? {
+                cut: true
+            } : {});
+        }
+        next = withPage(next, slot, {
+            read: kept,
             refused: null,
             title: clipLabel(read.title) || ((_next_pages_find = next.pages.find((page)=>page.slot === slot)) == null ? void 0 : _next_pages_find.title) || ''
         });
@@ -440,12 +707,13 @@ const withWave = (state, index, patch)=>_extends({}, state, {
         photos
     });
 }
-/** A page is in: its text is no longer kept. */ export function aiSiteImportPageSettled(state, slot) {
+/** A page is in: its text and its blocks are no longer kept. */ export function aiSiteImportPageSettled(state, slot) {
     const page = state.pages.find((entry)=>entry.slot === slot);
     if (!(page == null ? void 0 : page.read)) return state;
     return withPage(state, slot, {
         read: _extends({}, page.read, {
-            text: ''
+            text: '',
+            blocks: []
         })
     });
 }
@@ -453,8 +721,9 @@ const withWave = (state, index, patch)=>_extends({}, state, {
     var _ref;
     var _state_waves_find;
     const later = state.waves.some((wave)=>wave.index > index && wave.status === 'waiting');
+    // Start every remaining wave: no review between waves, the last one's close is the stop.
     let next = withWave(state, index, {
-        status: later ? 'review' : 'done',
+        status: later && !state.autoContinue ? 'review' : 'done',
         finishedAt: at
     });
     // What the wave read is no longer needed by any model answer.
@@ -467,7 +736,9 @@ const withWave = (state, index, patch)=>_extends({}, state, {
         pages: state.pages.map((page)=>page.read ? _extends({}, page, {
                 read: _extends({}, page.read, {
                     text: '',
-                    images: []
+                    images: [],
+                    blocks: [],
+                    forms: []
                 })
             }) : page)
     });
@@ -475,44 +746,103 @@ const withWave = (state, index, patch)=>_extends({}, state, {
 /*------------------------------------------
  * Credits
  *-----------------------------------------*/ /**
+ * How a page is built (AGL-4083), decided by code: every page of a collection
+ * (a blog, the legal pages, news) and any page longer than
+ * `AI_SITE_IMPORT_PROSE_WORDS` is copied as it is; the main pages are
+ * arranged by Aglyn AI, their words kept as written.
+ */ export function aiSiteImportPageBuild(page) {
+    return page.group !== 'main' || page.words > AI_SITE_IMPORT_PROSE_WORDS ? 'copied' : 'arranged';
+}
+const whole = (value)=>Math.max(0, Math.floor(Number.isFinite(value) ? value : 0));
+/**
  * What bringing in this many pages costs, quoted before it starts (AGL-3976):
  * each page an AI page as measured, its ceiling its two passes, and the
- * site's header and footer once when the wave builds them.
+ * site's header and footer once when the wave builds them. Pages the step has
+ * chosen a build for (AGL-4083) are priced by it: `arranged` at the
+ * arrangement answer and its listing, `copied` at the listing alone.
  */ export function aiSiteImportCreditRange(pages, options = {}) {
-    const count = Math.max(0, Math.floor(Number.isFinite(pages) ? pages : 0));
-    const { page, frame } = AI_MEASURED_IMPORT_CREDITS;
+    const count = whole(pages);
+    const arranged = whole(options.arranged);
+    const copied = whole(options.copied);
+    const measured = AI_MEASURED_IMPORT_CREDITS;
     return aiCreditRangeOrdered({
-        likely: count * page.median + (options.frame ? frame.median : 0),
-        p90: count * page.p90 + (options.frame ? frame.p90 : 0),
-        ceiling: count * AI_SITE_IMPORT_PAGE_CEILING_CREDITS + (options.frame ? AI_SITE_IMPORT_FRAME_CEILING_CREDITS : 0)
+        likely: count * measured.page.median + arranged * measured.arranged.median + copied * measured.copied.median + (options.frame ? measured.frame.median : 0),
+        p90: count * measured.page.p90 + arranged * measured.arranged.p90 + copied * measured.copied.p90 + (options.frame ? measured.frame.p90 : 0),
+        ceiling: (count + arranged) * AI_SITE_IMPORT_PAGE_CEILING_CREDITS + copied * AI_SITE_IMPORT_COPIED_CEILING_CREDITS + (options.frame ? AI_SITE_IMPORT_FRAME_CEILING_CREDITS : 0)
     });
 }
-/** A wave's quote: the pages it has still to bring in, and its header and footer while they are owed. */ export function aiSiteImportWaveCreditRange(state, wave, items) {
+/** What a wave still owes, by how its pages are built, and whether its header and footer are owed. */ export function aiSiteImportWaveOwed(state, wave, items) {
     const owed = (slot)=>items ? aiSiteImportItemOpen(items, slot) : true;
-    const pages = wave.slots.filter((slot)=>{
-        var _state_pages_find;
-        return owed(slot) && !((_state_pages_find = state.pages.find((page)=>page.slot === slot)) == null ? void 0 : _state_pages_find.refused);
-    }).length;
-    return aiSiteImportCreditRange(pages, {
+    const counts = {
+        pages: 0,
+        arranged: 0,
+        copied: 0,
         frame: wave.frame && !!state.frame && owed(AI_SITE_IMPORT_FRAME_SLOT)
+    };
+    for (const slot of wave.slots){
+        const page = state.pages.find((entry)=>entry.slot === slot);
+        if (!owed(slot) || (page == null ? void 0 : page.refused)) continue;
+        if ((page == null ? void 0 : page.build) === 'arranged') counts.arranged += 1;
+        else if ((page == null ? void 0 : page.build) === 'copied') counts.copied += 1;
+        else counts.pages += 1;
+    }
+    return counts;
+}
+/** A wave's quote: the pages it has still to bring in, and its header and footer while they are owed. */ export function aiSiteImportWaveCreditRange(state, wave, items) {
+    const owed = aiSiteImportWaveOwed(state, wave, items);
+    return aiSiteImportCreditRange(owed.pages, {
+        frame: owed.frame,
+        arranged: owed.arranged,
+        copied: owed.copied
     });
 }
+/** Every wave still waiting, quoted at once: "Start every remaining wave" (AGL-4083). */ export function aiSiteImportRemainingCreditRange(state, items) {
+    const total = {
+        pages: 0,
+        arranged: 0,
+        copied: 0,
+        frame: false
+    };
+    for (const wave of state.waves){
+        if (wave.status !== 'waiting') continue;
+        const owed = aiSiteImportWaveOwed(state, wave, items);
+        total.pages += owed.pages;
+        total.arranged += owed.arranged;
+        total.copied += owed.copied;
+        total.frame = total.frame || owed.frame;
+    }
+    return aiSiteImportCreditRange(total.pages, {
+        frame: total.frame,
+        arranged: total.arranged,
+        copied: total.copied
+    });
+}
+/** "About 95 credits (up to 600): 7 pages arranged by Aglyn AI, 3 copied as they are"; the build is named once the step has chosen. */ export function aiSiteImportQuoteText(range, owed) {
+    const parts = [];
+    if (owed.arranged) parts.push(`${aiSiteImportPagesText(owed.arranged)} arranged by ${PLATFORM_BRAND_NAME} AI`);
+    if (owed.copied) parts.push(`${owed.arranged ? owed.copied.toLocaleString('en-US') : aiSiteImportPagesText(owed.copied)} copied as ${owed.copied === 1 ? 'it is' : 'they are'}`);
+    return parts.length ? `${aiCreditRangeText(range)}: ${parts.join(', ')}` : aiCreditRangeText(range);
+}
 /*------------------------------------------
  * Words
  *-----------------------------------------*/ /** Why a page is not brought in, as a person reads it. */ export const AI_SITE_IMPORT_LEFT_OUT_COPY = {
     'plan-limit': 'Past your plan’s page count. Upgrade for more pages, then bring the rest in.',
     robots: 'The site’s robots.txt asks AglynBot not to read it.',
     'login-form': 'A sign-in page. Pages that ask for a password are never brought in.',
-    empty: 'Nothing to bring in: the page has no text or photos.',
+    empty: 'No text could be read from this page. If your site is built in a way that needs a browser to show its words, it cannot be brought in yet.',
     unreadable: 'It could not be read just now.',
     blocked: 'The site’s bot protection or firewall turned AglynBot away. Allow AglynBot in your site’s bot or firewall settings, then bring the site in again.',
     'too-large': 'Larger than a page can be to bring in.',
-    'off-site': 'It sends visitors to another website.'
+    'off-site': 'It sends visitors to another website.',
+    'not-reached': 'Not reached: the read stopped before it got to this page.'
 };
 /** "3 pages" / "1 page". */ export function aiSiteImportPagesText(count) {
     return `${count.toLocaleString('en-US')} ${count === 1 ? 'page' : 'pages'}`;
 }
 /** The review a finished wave parks on, or the first wave's go-ahead once the site is read. */ export function aiSiteImportWaveMessage(state, items) {
+    var _aiSiteImportReadVerdictMessage;
+    // A read that stopped at its home page says why, and what to do.
+    if (aiSiteImportReadStops(state.verdict)) return (_aiSiteImportReadVerdictMessage = aiSiteImportReadVerdictMessage(state.verdict, state.domain)) != null ? _aiSiteImportReadVerdictMessage : '';
     const next = aiSiteImportNextWave(state);
     const reviewing = aiSiteImportReviewWave(state);
     if (!reviewing) {