agent-sanitizer 2.61.0 → 2.61.2

package/src/unicode-properties.mjs

@@ -0,0 +1,365 @@
+/**
+ * GENERATED by scripts/gen-unicode-properties.mjs from the pinned UCD 17.0.0 — DO NOT EDIT.
+ *
+ * The Unicode properties src/invisible.mjs's carve-out reads, PINNED so the
+ * carve-out answers the same on every Node whatever its ICU's Unicode version.
+ * Regenerate with `node scripts/gen-unicode-properties.mjs`;
+ * test/unicode-properties.test.mjs fails if this drifts from the pinned package.
+ */
+
+export const UNICODE_VERSION = "17.0.0";
+
+// GENERATED UCD data, not hand-written logic: each table is the sorted, merged
+// [start, end] spans of one property. Pinned by the SSOT round-trip in
+// test/unicode-properties.test.mjs, so Stryker skips the numeric literals.
+// Stryker disable all
+/** @type {ReadonlyArray<readonly [number, number]>} */
+export const EXTENDED_PICTOGRAPHIC = [
+  [0xa9, 0xa9],
+  [0xae, 0xae],
+  [0x203c, 0x203c],
+  [0x2049, 0x2049],
+  [0x2122, 0x2122],
+  [0x2139, 0x2139],
+  [0x2194, 0x2199],
+  [0x21a9, 0x21aa],
+  [0x231a, 0x231b],
+  [0x2328, 0x2328],
+  [0x23cf, 0x23cf],
+  [0x23e9, 0x23f3],
+  [0x23f8, 0x23fa],
+  [0x24c2, 0x24c2],
+  [0x25aa, 0x25ab],
+  [0x25b6, 0x25b6],
+  [0x25c0, 0x25c0],
+  [0x25fb, 0x25fe],
+  [0x2600, 0x2604],
+  [0x260e, 0x260e],
+  [0x2611, 0x2611],
+  [0x2614, 0x2615],
+  [0x2618, 0x2618],
+  [0x261d, 0x261d],
+  [0x2620, 0x2620],
+  [0x2622, 0x2623],
+  [0x2626, 0x2626],
+  [0x262a, 0x262a],
+  [0x262e, 0x262f],
+  [0x2638, 0x263a],
+  [0x2640, 0x2640],
+  [0x2642, 0x2642],
+  [0x2648, 0x2653],
+  [0x265f, 0x2660],
+  [0x2663, 0x2663],
+  [0x2665, 0x2666],
+  [0x2668, 0x2668],
+  [0x267b, 0x267b],
+  [0x267e, 0x267f],
+  [0x2692, 0x2697],
+  [0x2699, 0x2699],
+  [0x269b, 0x269c],
+  [0x26a0, 0x26a1],
+  [0x26a7, 0x26a7],
+  [0x26aa, 0x26ab],
+  [0x26b0, 0x26b1],
+  [0x26bd, 0x26be],
+  [0x26c4, 0x26c5],
+  [0x26c8, 0x26c8],
+  [0x26ce, 0x26cf],
+  [0x26d1, 0x26d1],
+  [0x26d3, 0x26d4],
+  [0x26e9, 0x26ea],
+  [0x26f0, 0x26f5],
+  [0x26f7, 0x26fa],
+  [0x26fd, 0x26fd],
+  [0x2702, 0x2702],
+  [0x2705, 0x2705],
+  [0x2708, 0x270d],
+  [0x270f, 0x270f],
+  [0x2712, 0x2712],
+  [0x2714, 0x2714],
+  [0x2716, 0x2716],
+  [0x271d, 0x271d],
+  [0x2721, 0x2721],
+  [0x2728, 0x2728],
+  [0x2733, 0x2734],
+  [0x2744, 0x2744],
+  [0x2747, 0x2747],
+  [0x274c, 0x274c],
+  [0x274e, 0x274e],
+  [0x2753, 0x2755],
+  [0x2757, 0x2757],
+  [0x2763, 0x2764],
+  [0x2795, 0x2797],
+  [0x27a1, 0x27a1],
+  [0x27b0, 0x27b0],
+  [0x27bf, 0x27bf],
+  [0x2934, 0x2935],
+  [0x2b05, 0x2b07],
+  [0x2b1b, 0x2b1c],
+  [0x2b50, 0x2b50],
+  [0x2b55, 0x2b55],
+  [0x3030, 0x3030],
+  [0x303d, 0x303d],
+  [0x3297, 0x3297],
+  [0x3299, 0x3299],
+  [0x1f004, 0x1f004],
+  [0x1f02c, 0x1f02f],
+  [0x1f094, 0x1f09f],
+  [0x1f0af, 0x1f0b0],
+  [0x1f0c0, 0x1f0c0],
+  [0x1f0cf, 0x1f0d0],
+  [0x1f0f6, 0x1f0ff],
+  [0x1f170, 0x1f171],
+  [0x1f17e, 0x1f17f],
+  [0x1f18e, 0x1f18e],
+  [0x1f191, 0x1f19a],
+  [0x1f1ae, 0x1f1e5],
+  [0x1f201, 0x1f20f],
+  [0x1f21a, 0x1f21a],
+  [0x1f22f, 0x1f22f],
+  [0x1f232, 0x1f23a],
+  [0x1f23c, 0x1f23f],
+  [0x1f249, 0x1f25f],
+  [0x1f266, 0x1f321],
+  [0x1f324, 0x1f393],
+  [0x1f396, 0x1f397],
+  [0x1f399, 0x1f39b],
+  [0x1f39e, 0x1f3f0],
+  [0x1f3f3, 0x1f3f5],
+  [0x1f3f7, 0x1f3fa],
+  [0x1f400, 0x1f4fd],
+  [0x1f4ff, 0x1f53d],
+  [0x1f549, 0x1f54e],
+  [0x1f550, 0x1f567],
+  [0x1f56f, 0x1f570],
+  [0x1f573, 0x1f57a],
+  [0x1f587, 0x1f587],
+  [0x1f58a, 0x1f58d],
+  [0x1f590, 0x1f590],
+  [0x1f595, 0x1f596],
+  [0x1f5a4, 0x1f5a5],
+  [0x1f5a8, 0x1f5a8],
+  [0x1f5b1, 0x1f5b2],
+  [0x1f5bc, 0x1f5bc],
+  [0x1f5c2, 0x1f5c4],
+  [0x1f5d1, 0x1f5d3],
+  [0x1f5dc, 0x1f5de],
+  [0x1f5e1, 0x1f5e1],
+  [0x1f5e3, 0x1f5e3],
+  [0x1f5e8, 0x1f5e8],
+  [0x1f5ef, 0x1f5ef],
+  [0x1f5f3, 0x1f5f3],
+  [0x1f5fa, 0x1f64f],
+  [0x1f680, 0x1f6c5],
+  [0x1f6cb, 0x1f6d2],
+  [0x1f6d5, 0x1f6e5],
+  [0x1f6e9, 0x1f6e9],
+  [0x1f6eb, 0x1f6f0],
+  [0x1f6f3, 0x1f6ff],
+  [0x1f7da, 0x1f7ff],
+  [0x1f80c, 0x1f80f],
+  [0x1f848, 0x1f84f],
+  [0x1f85a, 0x1f85f],
+  [0x1f888, 0x1f88f],
+  [0x1f8ae, 0x1f8af],
+  [0x1f8bc, 0x1f8bf],
+  [0x1f8c2, 0x1f8cf],
+  [0x1f8d9, 0x1f8ff],
+  [0x1f90c, 0x1f93a],
+  [0x1f93c, 0x1f945],
+  [0x1f947, 0x1f9ff],
+  [0x1fa58, 0x1fa5f],
+  [0x1fa6e, 0x1faff],
+  [0x1fc00, 0x1fffd],
+];
+
+/** @type {ReadonlyArray<readonly [number, number]>} */
+export const EMOJI_MODIFIER = [
+  [0x1f3fb, 0x1f3ff],
+];
+
+/** @type {ReadonlyArray<readonly [number, number]>} */
+export const EMOJI_TEXT_DEFAULT = [
+  [0x23, 0x23],
+  [0x2a, 0x2a],
+  [0x30, 0x39],
+  [0xa9, 0xa9],
+  [0xae, 0xae],
+  [0x203c, 0x203c],
+  [0x2049, 0x2049],
+  [0x2122, 0x2122],
+  [0x2139, 0x2139],
+  [0x2194, 0x2199],
+  [0x21a9, 0x21aa],
+  [0x2328, 0x2328],
+  [0x23cf, 0x23cf],
+  [0x23ed, 0x23ef],
+  [0x23f1, 0x23f2],
+  [0x23f8, 0x23fa],
+  [0x24c2, 0x24c2],
+  [0x25aa, 0x25ab],
+  [0x25b6, 0x25b6],
+  [0x25c0, 0x25c0],
+  [0x25fb, 0x25fc],
+  [0x2600, 0x2604],
+  [0x260e, 0x260e],
+  [0x2611, 0x2611],
+  [0x2618, 0x2618],
+  [0x261d, 0x261d],
+  [0x2620, 0x2620],
+  [0x2622, 0x2623],
+  [0x2626, 0x2626],
+  [0x262a, 0x262a],
+  [0x262e, 0x262f],
+  [0x2638, 0x263a],
+  [0x2640, 0x2640],
+  [0x2642, 0x2642],
+  [0x265f, 0x2660],
+  [0x2663, 0x2663],
+  [0x2665, 0x2666],
+  [0x2668, 0x2668],
+  [0x267b, 0x267b],
+  [0x267e, 0x267e],
+  [0x2692, 0x2692],
+  [0x2694, 0x2697],
+  [0x2699, 0x2699],
+  [0x269b, 0x269c],
+  [0x26a0, 0x26a0],
+  [0x26a7, 0x26a7],
+  [0x26b0, 0x26b1],
+  [0x26c8, 0x26c8],
+  [0x26cf, 0x26cf],
+  [0x26d1, 0x26d1],
+  [0x26d3, 0x26d3],
+  [0x26e9, 0x26e9],
+  [0x26f0, 0x26f1],
+  [0x26f4, 0x26f4],
+  [0x26f7, 0x26f9],
+  [0x2702, 0x2702],
+  [0x2708, 0x2709],
+  [0x270c, 0x270d],
+  [0x270f, 0x270f],
+  [0x2712, 0x2712],
+  [0x2714, 0x2714],
+  [0x2716, 0x2716],
+  [0x271d, 0x271d],
+  [0x2721, 0x2721],
+  [0x2733, 0x2734],
+  [0x2744, 0x2744],
+  [0x2747, 0x2747],
+  [0x2763, 0x2764],
+  [0x27a1, 0x27a1],
+  [0x2934, 0x2935],
+  [0x2b05, 0x2b07],
+  [0x3030, 0x3030],
+  [0x303d, 0x303d],
+  [0x3297, 0x3297],
+  [0x3299, 0x3299],
+  [0x1f170, 0x1f171],
+  [0x1f17e, 0x1f17f],
+  [0x1f202, 0x1f202],
+  [0x1f237, 0x1f237],
+  [0x1f321, 0x1f321],
+  [0x1f324, 0x1f32c],
+  [0x1f336, 0x1f336],
+  [0x1f37d, 0x1f37d],
+  [0x1f396, 0x1f397],
+  [0x1f399, 0x1f39b],
+  [0x1f39e, 0x1f39f],
+  [0x1f3cb, 0x1f3ce],
+  [0x1f3d4, 0x1f3df],
+  [0x1f3f3, 0x1f3f3],
+  [0x1f3f5, 0x1f3f5],
+  [0x1f3f7, 0x1f3f7],
+  [0x1f43f, 0x1f43f],
+  [0x1f441, 0x1f441],
+  [0x1f4fd, 0x1f4fd],
+  [0x1f549, 0x1f54a],
+  [0x1f56f, 0x1f570],
+  [0x1f573, 0x1f579],
+  [0x1f587, 0x1f587],
+  [0x1f58a, 0x1f58d],
+  [0x1f590, 0x1f590],
+  [0x1f5a5, 0x1f5a5],
+  [0x1f5a8, 0x1f5a8],
+  [0x1f5b1, 0x1f5b2],
+  [0x1f5bc, 0x1f5bc],
+  [0x1f5c2, 0x1f5c4],
+  [0x1f5d1, 0x1f5d3],
+  [0x1f5dc, 0x1f5de],
+  [0x1f5e1, 0x1f5e1],
+  [0x1f5e3, 0x1f5e3],
+  [0x1f5e8, 0x1f5e8],
+  [0x1f5ef, 0x1f5ef],
+  [0x1f5f3, 0x1f5f3],
+  [0x1f5fa, 0x1f5fa],
+  [0x1f6cb, 0x1f6cb],
+  [0x1f6cd, 0x1f6cf],
+  [0x1f6e0, 0x1f6e5],
+  [0x1f6e9, 0x1f6e9],
+  [0x1f6f0, 0x1f6f0],
+  [0x1f6f3, 0x1f6f3],
+];
+
+/** @type {ReadonlyArray<readonly [number, number]>} */
+export const UNIFIED_IDEOGRAPH = [
+  [0x3400, 0x4dbf],
+  [0x4e00, 0x9fff],
+  [0xfa0e, 0xfa0f],
+  [0xfa11, 0xfa11],
+  [0xfa13, 0xfa14],
+  [0xfa1f, 0xfa1f],
+  [0xfa21, 0xfa21],
+  [0xfa23, 0xfa24],
+  [0xfa27, 0xfa29],
+  [0x20000, 0x2a6df],
+  [0x2a700, 0x2b81d],
+  [0x2b820, 0x2cead],
+  [0x2ceb0, 0x2ebe0],
+  [0x2ebf0, 0x2ee5d],
+  [0x30000, 0x3134a],
+  [0x31350, 0x33479],
+];
+
+/** @type {ReadonlyArray<readonly [number, number]>} */
+export const SCRIPT_BRAILLE = [
+  [0x2800, 0x28ff],
+];
+
+/** @type {ReadonlyArray<readonly [number, number]>} */
+export const SCRIPT_HANGUL = [
+  [0x1100, 0x11ff],
+  [0x302e, 0x302f],
+  [0x3131, 0x318e],
+  [0x3200, 0x321e],
+  [0x3260, 0x327e],
+  [0xa960, 0xa97c],
+  [0xac00, 0xd7a3],
+  [0xd7b0, 0xd7c6],
+  [0xd7cb, 0xd7fb],
+  [0xffa0, 0xffbe],
+  [0xffc2, 0xffc7],
+  [0xffca, 0xffcf],
+  [0xffd2, 0xffd7],
+  [0xffda, 0xffdc],
+];
+// Stryker restore all
+
+/**
+ * True when `cp` lies in one of `spans` (sorted, non-overlapping). A negative
+ * `cp` — the "no neighbour" sentinel — lies in none.
+ * @param {ReadonlyArray<readonly [number, number]>} spans
+ * @param {number} cp
+ * @returns {boolean}
+ */
+export function inSpans(spans, cp) {
+  let lo = 0;
+  let hi = spans.length - 1;
+  while (lo <= hi) {
+    const mid = (lo + hi) >> 1;
+    if (cp < spans[mid][0]) hi = mid - 1;
+    else if (cp > spans[mid][1]) lo = mid + 1;
+    else return true;
+  }
+  return false;
+}

package/types/ansi.d.mts

@@ -49,15 +49,12 @@ export function isCancelledOrphan(text: string, token: AnsiToken): boolean;
  */
 export function scanAnsi(text: string): AnsiToken[];
 /**
- * The CONCEAL state after the SGR token `token` is applied to a terminal already
- * in state `concealed` — the state in which a terminal renders what follows as
- * blank while its bytes stay readable to anything reading the file.
+ * The rendition after the SGR token `token` is applied to a terminal already in
+ * `state`.
  *
- * A TRANSITION, not a property of the token: conceal is terminal state that
+ * A TRANSITION, not a property of the token: rendition is terminal state that
  * outlives the sequence that set it, so `ESC[8m` followed by `ESC[31m` is still
- * concealed — a per-token predicate would read the second token as "no conceal
- * here" and lose the state. Only `8` (set), `28` (reveal) and `0` (reset all)
- * move it, and the token's parameters are applied in order, so a later reveal or
+ * concealed. The token's parameters are applied in order, so a later reveal or
  * reset in the SAME token cancels an earlier `8`: `ESC[8;28;31m` renders red and
  * visible.
  *
@@ -65,18 +62,26 @@ export function scanAnsi(text: string): AnsiToken[];
  * the digit. `38`/`48`/`58` take their colour arguments from the parameters that
  * FOLLOW them in the semicolon form, so `ESC[38;5;8m` is bright-black foreground
  * and its `8` is a palette index; a parameter carrying its arguments as ITU T.416
- * sub-parameters (`38:5:8`) is self-contained, so only its head counts and
- * nothing after it is consumed. A colour-space selector T.416 gives no argument
- * count for is malformed, and a terminal that ignores the colour form still
- * applies what follows it, so the scan CONTINUES from the next parameter rather
- * than consuming arguments it cannot size — reading one parameter too many costs
- * a token shape no emitter produces, while stopping there would hand
- * `ESC[38;9;8m` a pass.
+ * sub-parameters (`38:5:8`) is self-contained, so nothing after it is consumed.
+ * A colour-space selector T.416 gives no argument count for is malformed, and a
+ * terminal that ignores the colour form still applies what follows it, so the
+ * scan CONTINUES from the next parameter rather than consuming arguments it
+ * cannot size — stopping there would hand `ESC[38;9;8m` a pass.
  * @param {string} token  a token {@link scanAnsi} classified {@link TOKEN_KIND.SGR}
- * @param {boolean} concealed  the state before this token
+ * @param {Readonly<SgrRendition>} state  the rendition before this token
+ * @returns {SgrRendition}
+ */
+export function sgrRendition(token: string, state: Readonly<SgrRendition>): SgrRendition;
+/**
+ * True when text drawn in `rendition` is invisible to a human because its
+ * foreground IS its background (`ESC[30;40m`, `ESC[38;5;7;48;5;7m`). Only an
+ * unambiguous match counts: a default colour is unknown, so it matches nothing,
+ * and BOLD over a basic colour is drawn brighter by many terminals, so
+ * `ESC[1;30;40m` is not called hidden — the precision direction.
+ * @param {Readonly<SgrRendition>} rendition
  * @returns {boolean}
  */
-export function sgrConcealState(token: string, concealed: boolean): boolean;
+export function foregroundMatchesBackground({ bold, fg, bg }: Readonly<SgrRendition>): boolean;
 /**
  * The ONE ANSI grammar: the raw control-introducer charset and the tokenizer
  * every consumer scans with.
@@ -170,6 +175,17 @@ export const TOKEN_KIND: Readonly<{
      */
     ORPHAN_C1: "orphan-c1-introducer";
 }>;
+/**
+ * What a terminal's graphic rendition looks like after an SGR token: whether
+ * CONCEAL is on, whether BOLD is (which many terminals render as a brighter
+ * foreground), and the foreground/background colour as a comparable key — `i<n>`
+ * for a palette index (`31` and `38;5;1` are both `i1`), `rgb:<r>,<g>,<b>` for
+ * direct colour, `null` for the default or a colour this reader cannot size.
+ * @typedef {{ concealed: boolean, bold: boolean, fg: string | null, bg: string | null }} SgrRendition
+ */
+/** The rendition of a terminal no SGR has touched (and of `ESC[0m`).
+ * @type {Readonly<SgrRendition>} */
+export const DEFAULT_RENDITION: Readonly<SgrRendition>;
 export type AnsiToken = {
     /**
      * Index of the introducer.
@@ -184,3 +200,16 @@ export type AnsiToken = {
      */
     kind: string;
 };
+/**
+ * What a terminal's graphic rendition looks like after an SGR token: whether
+ * CONCEAL is on, whether BOLD is (which many terminals render as a brighter
+ * foreground), and the foreground/background colour as a comparable key — `i<n>`
+ * for a palette index (`31` and `38;5;1` are both `i1`), `rgb:<r>,<g>,<b>` for
+ * direct colour, `null` for the default or a colour this reader cannot size.
+ */
+export type SgrRendition = {
+    concealed: boolean;
+    bold: boolean;
+    fg: string | null;
+    bg: string | null;
+};

package/types/cf-charset.d.mts

@@ -1,14 +1,14 @@
 /**
- * GENERATED by scripts/gen-invisible-charset.mjs from Node's Unicode data
- * (\p{Cf}, Unicode 17.0) — DO NOT EDIT.
+ * GENERATED by scripts/gen-invisible-charset.mjs from the pinned UCD
+ * (General_Category=Cf, Unicode 17.0.0) — DO NOT EDIT.
  *
  * The general-category Cf code points, PINNED at generation time. invisible.mjs
- * strips exactly this set instead of testing \p{Cf} live, and the Python port
+ * strips exactly this set instead of asking the runtime for Cf, and the Python port
  * reads the SAME set from data/invisible-charset.json's `cf_codepoints`, so both
  * layers strip an identical Cf set regardless of each runtime's own Unicode
  * version. Regenerate with `node scripts/gen-invisible-charset.mjs`;
- * test/invisible-charset.test.mjs fails if this drifts from Node's \p{Cf}.
+ * test/invisible-charset.test.mjs fails if this drifts from the pinned UCD.
  */
-export const UNICODE_VERSION: "17.0";
+export const UNICODE_VERSION: "17.0.0";
 /** @type {readonly number[]} Sorted ascending. */
 export const CF_CODEPOINTS: readonly number[];

package/types/index.d.mts

@@ -58,6 +58,6 @@ export function sanitize(text: string, options?: {
         original: string;
     }>;
 }>;
-export { applyLayer1, applyLayer1WellFormed, isBenignAnsi, isBenignAnsiKinds, normalizeLoneSurrogates, stripAnsiFully, sgrCarriesPayload, SGR_RUN_THRESHOLD, LONE_SURROGATE_RE } from "./layer1.mjs";
+export { applyLayer1, applyLayer1WellFormed, escapesPreservable, isBenignAnsi, isBenignAnsiKinds, normalizeLoneSurrogates, stripAnsiFully, stripTerminalFully, sgrCarriesPayload, SGR_RUN_THRESHOLD, LONE_SURROGATE_RE } from "./layer1.mjs";
 export { stripInvisible, stripInvisibleWithReport, isSgrOnly, STRIP, SGR_RE, CHECKS, CATEGORY, CATEGORY_LABELS, LINGUISTIC_SCRIPTS, VS, BLANK_NON_CF, LONG_RUN_RE, LONG_RUN_THRESHOLD, SCATTERED_THRESHOLD, findLongRuns, hasLongRun } from "./invisible.mjs";
 export { HTML_TAG_PRESENT, MD_LINK_HINT, SECRET_HINT, SECRET_HINT_EXT, matchesSecretHint } from "./gates.mjs";

package/types/instructions.d.mts

@@ -14,9 +14,13 @@ export function decodeRun(run: string): {
     decoded: string;
 };
 /**
- * Scan a file's text for hidden-Unicode injection. Reports each long invisible
+ * Scan a file's text for hidden-Unicode injection. Reports each payload long
  * run (with its decoded payload) plus a single scattered-chars finding when the
- * non-run invisible count crosses the threshold-evasion floor.
+ * effective invisible count outside those runs crosses the threshold-evasion
+ * floor.
+ *
+ * Whether there is anything to report is {@link payloadVerdict}'s call, the
+ * one every other surface makes.
  * @param {string} content
  * @returns {Array<{ line: number | null, charCount: number, method: string, decoded: string }>}
  *   `line` is the 1-based line of a long-run finding, or `null` for the

package/types/invisible.d.mts

@@ -68,11 +68,13 @@ export function describeStripped(invisFound: string[], deAnsi: string): string;
  */
 export function countPayloadInvisible(text: string): number;
 /**
- * The `text` with every carve-out-PRESERVABLE invisible (joiners/selectors/tags/
- * blank fillers doing real rendering work) replaced by a space, leaving only the
- * PAYLOAD invisibles in place. The LONG_RUN injection probe runs over this so a
- * legitimate emoji/flag/variation sequence never trips the "possible injection
- * payload" marker (alert fatigue), while a genuine hidden run still surfaces.
+ * The `text` with every visible character and every carve-out-PRESERVABLE
+ * invisible (joiners/selectors/tags/blank fillers doing real rendering work)
+ * replaced by spaces — one per UTF-16 unit, so the view is index-aligned with
+ * `text` — leaving only the PAYLOAD invisibles in place. The LONG_RUN injection
+ * probe runs over this so a legitimate emoji/flag/variation sequence never trips
+ * the "possible injection payload" marker (alert fatigue), while a genuine
+ * hidden run still surfaces.
  * @param {string} text
  * @returns {string}
  */
@@ -143,6 +145,25 @@ export function countEffectiveInvisible(text: string): number;
  * @returns {boolean}
  */
 export function isIncidentalInvisible(text: string): boolean;
+/**
+ * THE decision whether `text` carries a payload-capable invisible channel: a
+ * payload LONG RUN, or {@link SCATTERED_THRESHOLD} or more effective payload
+ * invisibles. Every gate reads this one verdict, so a Persian paragraph cannot
+ * be a payload to one and clean to another. It stops at the first run, so an
+ * 8 MB field of short runs costs one record.
+ * @param {string} text  the text a consumer acts on, ANSI-stripped where its
+ *   surface can carry escapes (a sequence can hide invisibles)
+ * @returns {{ firstRun: { index: number, text: string, charCount: number } | null, invisibleCount: number, payloadCapable: boolean }}
+ */
+export function payloadVerdict(text: string): {
+    firstRun: {
+        index: number;
+        text: string;
+        charCount: number;
+    } | null;
+    invisibleCount: number;
+    payloadCapable: boolean;
+};
 /**
  * Strip payload-capable invisible chars and report which categories were
  * removed. A single leading U+FEFF (BOM) is preserved as a legitimate marker;
@@ -155,11 +176,17 @@ export function isIncidentalInvisible(text: string): boolean;
  * decide whether a leading BOM is genuinely leading: an interior BOM that an
  * ANSI-strip left at index 0 of `text` (e.g. `ESC[m + interior U+FEFF`) must NOT be treated as a
  * legitimate leading marker. Defaults to `text` for the common single-arg call.
+ *
+ * `carveOut: false` strips EVERY payload-capable invisible, preserved ones
+ * included — the monotone fallback for a composition that fails to converge.
  * @param {string} text
  * @param {string} [originalText]
+ * @param {{ carveOut?: boolean }} [options]
  * @returns {{ cleaned: string, found: string[] }}
  */
-export function stripInvisibleWithReport(text: string, originalText?: string): {
+export function stripInvisibleWithReport(text: string, originalText?: string, { carveOut }?: {
+    carveOut?: boolean;
+}): {
     cleaned: string;
     found: string[];
 };
@@ -215,3 +242,31 @@ export const TOTAL_PRESERVED_BLANK_BUDGET: 16;
 export const PRESERVED_BLANK_PER_ANCHOR: 2;
 export const LINGUISTIC_SCRIPTS: string[];
 export { BRAHMIC_CONSONANT_RANGES } from "./joining-type.mjs";
+export type CarveEntry = {
+    body: string;
+    cps: Int32Array;
+    analysis: ReturnType<typeof analyzeCarve>;
+    strip: {
+        cleaned: string;
+        found: string[];
+        removed: number;
+    } | null;
+    removed: number | null;
+};
+/**
+ * Per-invisible carve-out analysis, shared by carveStrip and
+ * countPayloadInvisible: for each code point, its CODE_* class (CODE_VISIBLE
+ * when visible) and its preserve KIND_* (KIND_NONE when the char is payload).
+ * Everything invisible that is NOT preserve-eligible is payload; the scatter
+ * floor counts only that, so meaningful joiners/selectors never push honest
+ * prose over the threshold.
+ * @param {Int32Array} cps
+ * @returns {{ codes: Uint8Array, kind: Uint8Array, invisible: Int32Array, payloadInvis: number, visibleLen: number }}
+ */
+declare function analyzeCarve(cps: Int32Array): {
+    codes: Uint8Array;
+    kind: Uint8Array;
+    invisible: Int32Array;
+    payloadInvis: number;
+    visibleLen: number;
+};

package/types/joining-type.d.mts

@@ -15,13 +15,16 @@ export function isVirama(cp: number): boolean;
 /**
  * True when `cp` is a Brahmic consonant — the only base a virama attaches to,
  * and therefore the only base after which a ZWJ/ZWNJ is a real conjunct
- * request rather than a zero-width payload.
+ * request rather than a zero-width payload. A binary search, like
+ * `lookup`: the spans are sorted and disjoint, and the carve analysis asks
+ * this of every visible code point of a long document, where a scan of all the
+ * spans per code point was the dominant cost.
  * @param {number} cp
  * @returns {boolean}
  */
 export function isBrahmicConsonant(cp: number): boolean;
 /**
- * GENERATED by scripts/gen-joining-type.mjs from ucd-full@17.0.0 — DO NOT EDIT.
+ * GENERATED by scripts/gen-joining-type.mjs from the pinned UCD 17.0.0 — DO NOT EDIT.
  *
  * Unicode Joining_Type, Indic virama and Brahmic consonant range tables backing
  * the ZWNJ/ZWJ carve-out in invisible.mjs. Regenerate with `pnpm gen:joining-type`;