text-sanctifier 1.3.3 → 1.3.5

package/src/sanctifyText.js

@@ -1,506 +1,563 @@
-// src/sanctifyText.js
-
-import { CONTROL_CHARS_REGEX, INVISIBLE_TRASH_REGEX } from './categories.generated.js';
-export { CONTROL_CHARS_REGEX, INVISIBLE_TRASH_REGEX };
-
-
-/**
- * @typedef {Object} SanctifyOptions
- * @property {boolean} [purgeInvisibleChars]
- * @property {boolean} [purgeEmojis]
- * @property {boolean} [nukeControls]
- * @property {boolean} [keyboardOnlyFilter]
- * @property {boolean} [normalizeNewlines]
- * @property {boolean} [trimSpacesAroundNewlines]
- * @property {boolean} [collapseNewLines]
- * @property {boolean} [preserveParagraphs]
- * @property {boolean} [collapseSpaces]
- * @property {boolean} [finalTrim]
- * @property {number} [maxLength]
- * @property {'throw'|'truncate'|'noop'} [onMaxLength]
- */
-
-
-/**
- * Summons a customized sanctifier function with pre-bound booleans.
- *
- * Accepts full flag names and returns a text-cleaning function.
- *
- * @param {Object} [defaultOptions={}]
- * @param {boolean} [defaultOptions.purgeInvisibleChars]
- * @param {boolean} [defaultOptions.purgeEmojis]
- * @param {boolean} [defaultOptions.nukeControls]
- * @param {boolean} [defaultOptions.keyboardOnlyFilter]
- * @param {boolean} [defaultOptions.normalizeNewlines]
- * @param {boolean} [defaultOptions.trimSpacesAroundNewlines]
- * @param {boolean} [defaultOptions.collapseNewLines]
- * @param {boolean} [defaultOptions.preserveParagraphs]
- * @param {boolean} [defaultOptions.collapseSpaces]
- * @param {boolean} [defaultOptions.finalTrim]
- * @param {number} [defaultOptions.maxLength=Infinity] - Hard input length cap (UTF-16 code units).
- * @param {'throw' | 'truncate' | 'noop'} [defaultOptions.onMaxLength='throw'] - Behavior when input exceeds maxLength.
- * @returns {(text: string) => string}
- */
-export function summonSanctifier(defaultOptions = {}) {
-    const purgeInvisibleChars = !!defaultOptions.purgeInvisibleChars;
-    const purgeEmojis = !!defaultOptions.purgeEmojis;
-    const nukeControls = !!defaultOptions.nukeControls;
-    const keyboardOnlyFilter = !!defaultOptions.keyboardOnlyFilter;
-    const normalizeNewlines = !!defaultOptions.normalizeNewlines;
-    const trimSpacesAroundNewlines = !!defaultOptions.trimSpacesAroundNewlines;
-    const collapseNewLines = !!defaultOptions.collapseNewLines;
-    const preserveParagraphs = !!defaultOptions.preserveParagraphs;
-    const collapseSpaces = !!defaultOptions.collapseSpaces;
-    const finalTrim = !!defaultOptions.finalTrim;
-
-    const maxLength =
-        Number.isFinite(defaultOptions.maxLength) && defaultOptions.maxLength >= 0
-            ? defaultOptions.maxLength
-            : Infinity;
-
-    const onMaxLength =
-        defaultOptions.onMaxLength === 'truncate' ||
-            defaultOptions.onMaxLength === 'noop' ||
-            defaultOptions.onMaxLength === 'throw'
-            ? defaultOptions.onMaxLength
-            : 'throw';
-
-    return (text) =>
-        sanctifyText(
-            text,
-            purgeInvisibleChars,
-            purgeEmojis,
-            nukeControls,
-            keyboardOnlyFilter,
-            normalizeNewlines,
-            trimSpacesAroundNewlines,
-            collapseNewLines,
-            preserveParagraphs,
-            collapseSpaces,
-            finalTrim,
-            maxLength,
-            onMaxLength
-        );
-}
-
-// --- Added Presets ---
-
-/**
- * Strict sanitizer:
- * - Purge emojis
- * - Collapse all newlines
- * - Collapse spaces
- * - Nuke control characters
- */
-/** @param {string} text @returns {string} */
-summonSanctifier.strict = text => sanctifyText(
-    text,
-    true,   // purgeInvisibleChars
-    true,   // purgeEmojis
-    true,   // nukeControls
-    false,  // keyboardOnlyFilter
-    true,   // normalizeNewlines
-    true,   // trimSpacesAroundNewlines
-    true,   // collapseNewLines
-    false,  // preserveParagraphs
-    true,   // collapseSpaces
-    true    // finalTrim
-);
-
-
-/**
- * Loose sanitizer:
- * - Collapse spaces
- * - Preserve paragraphs
- * - Normalize newlines
- */
-/** @param {string} text @returns {string} */
-summonSanctifier.loose = text => sanctifyText(
-    text,
-    false,  // purgeInvisibleChars
-    false,  // purgeEmojis
-    false,  // nukeControls
-    false,  // keyboardOnlyFilter
-    true,   // normalizeNewlines
-    true,   // trimSpacesAroundNewlines
-    true,   // collapseNewLines
-    true,   // preserveParagraphs
-    true,   // collapseSpaces
-    true    // finalTrim
-);
-
-
-/**
- * Keyboard-only (with emojis):
- * - Keeps emojis and printable ASCII
- * - Strips non-standard characters
- * - Normalizes typographic trash
- */
-/** @param {string} text @returns {string} */
-summonSanctifier.keyboardOnlyEmoji = text => sanctifyText(
-    text,
-    false,  // purgeInvisibleChars
-    false,  // purgeEmojis
-    false,  // nukeControls
-    true,   // keyboardOnlyFilter
-    true,   // normalizeNewlines
-    true,   // trimSpacesAroundNewlines
-    false,  // collapseNewLines
-    false,  // preserveParagraphs
-    false,  // collapseSpaces
-    true    // finalTrim
-);
-
-
-/**
- * Keyboard-only (strict):
- * - Removes emojis
- * - Collapses all whitespace
- * - Restricts to printable ASCII only
- */
-/** @param {string} text @returns {string} */
-summonSanctifier.keyboardOnly = text => sanctifyText(
-    text,
-    true,   // purgeInvisibleChars
-    true,   // purgeEmojis
-    true,   // nukeControls
-    true,   // keyboardOnlyFilter
-    true,   // normalizeNewlines
-    true,   // trimSpacesAroundNewlines
-    true,   // collapseNewLines
-    false,  // preserveParagraphs
-    true,   // collapseSpaces
-    true    // finalTrim
-);
-
-
-/**
- * Text Sanctifier
- *
- * Brutal text normalizer and invisible trash scrubber,
- * configurable to kill whatever ghosts you want dead.
- *
- * ⚠️ Note: This is plain-text normalization/filtering — not an HTML/XSS sanitizer.
- *
- * @param {string} text
- * @param {boolean} [purgeInvisibleChars=false] - Remove ZWSP, NBSP, bidi, tags, variation selectors, etc.
- * @param {boolean} [purgeEmojis=false] - Remove emoji characters entirely.
- * @param {boolean} [nukeControls=false] - Remove non-whitespace control characters.
- * @param {boolean} [keyboardOnlyFilter=false] - Keep printable ASCII + full emoji sequences only (drops other Unicode).
- * @param {boolean} [normalizeNewlines=false] - Convert all newlines to `\n`.
- * @param {boolean} [trimSpacesAroundNewlines=false] - Remove spaces/tabs around newlines.
- * @param {boolean} [collapseNewLines=false] - Collapse `\n` runs (optionally preserve paragraphs).
- * @param {boolean} [preserveParagraphs=false] - Preserve paragraph breaks when collapsing newlines.
- * @param {boolean} [collapseSpaces=false] - Collapse multiple spaces into one.
- * @param {boolean} [finalTrim=false] - `.trim()` the final output (head/tail).
- * @param {number} [maxLength=Infinity] - Hard input length cap (UTF-16 code units via `text.length`).
- * @param {'throw'|'truncate'|'noop'} [onMaxLength='throw'] - Behavior when `text.length` exceeds `maxLength`:
- *   - `'throw'`: throw a `RangeError`
- *   - `'truncate'`: slice to `maxLength` before processing
- *   - `'noop'`: return the original input unchanged (skips processing)
- * @returns {string}
- * @throws {TypeError} If `text` is not a string.
- * @throws {RangeError} If input exceeds `maxLength` and `onMaxLength` is `'throw'`.
- */
-export function sanctifyText(
-    text,
-    purgeInvisibleChars = false,
-    purgeEmojis = false,
-    nukeControls = false,
-    keyboardOnlyFilter = false,
-    normalizeNewlines = false,
-    trimSpacesAroundNewlines = false,
-    collapseNewLines = false,
-    preserveParagraphs = false,
-    collapseSpaces = false,
-    finalTrim = false,
-    maxLength = Infinity,
-    onMaxLength = 'throw'
-) {
-    if (typeof text !== 'string') {
-        throw new TypeError('sanctifyText expects a string input.');
-    }
-
-    if (onMaxLength !== 'truncate' && onMaxLength !== 'noop' && onMaxLength !== 'throw') {
-        onMaxLength = 'throw';
-    }
-
-    if (text.length > maxLength) {
-        switch (onMaxLength) {
-            case 'truncate':
-                text = text.slice(0, maxLength);
-                break;
-            case 'noop':
-                return text;
-            case 'throw':
-            default:
-                throw new RangeError(
-                    `sanctifyText input length ${text.length} exceeds maxLength ${maxLength}.`
-                );
-        }
-    }
-
-    let cleaned = text;
-
-    //  Purge invisible Unicode trash (zero-width, non-breaking, bidi junk, etc.)
-    if (purgeInvisibleChars) cleaned = purgeInvisibleTrash(cleaned);
-
-    // Remove emojis
-    if (purgeEmojis) cleaned = purgeEmojiCharacters(cleaned);
-
-    // Nuke control characters (excluding whitespace)
-    if (nukeControls) cleaned = purgeControlCharacters(cleaned);
-
-    // Keep only ASCII/emojis
-    if (keyboardOnlyFilter) cleaned = purgeNonKeyboardChars(cleaned, purgeEmojis);
-
-    // Normalize line endings to Unix style (\n)
-    if (normalizeNewlines) cleaned = normalizeNewlineChars(cleaned);
-
-    // Remove spaces/tabs around newlines
-    if (trimSpacesAroundNewlines) cleaned = trimSpacesAroundNewlineChars(cleaned);
-
-    // Collapse excessive newlines, Optionally preserve Paragraphs
-    if (collapseNewLines) cleaned = collapseMultipleNewLines(cleaned, preserveParagraphs);
-
-    // Collapse multiple spaces into a single space
-    if (collapseSpaces) cleaned = collapseExtraSpaces(cleaned);
-
-    // Final trim, return Sanctified Text
-    return finalTrim ? cleaned.trim() : cleaned;
-}
-
-
-// --- Micro helpers ---
-
-
-/**
- * Removes the invisible Unicode characters in INVISIBLE_TRASH_REGEX: non-breaking and
- * typographic spaces, zero-width characters, bidi controls, invisible format characters,
- * Hangul fillers, variation selectors (including VS16 in emoji), the BOM and tag characters.
- * The code points come from the table in scripts/categories.js.
- *
- * @param {string} text
- * @returns {string}
- */
-function purgeInvisibleTrash(text) {
-    return text.replace(INVISIBLE_TRASH_REGEX, '');
-}
-
-
-/**
- * Extracts and preserves complete emoji sequences from a string.
- *
- * Uses the global `EMOJI_REGEX` to match full emoji grapheme sequences,
- * including:
- *   - Extended pictographic characters
- *   - Zero-width joiner (ZWJ) sequences (e.g. 👨‍👩‍👧‍👦)
- *   - Variation Selector-16 (VS16) emoji presentation forms (e.g. ✌️, ‼️)
- *
- * Any non-emoji characters are ignored.
- *
- * Important:
- * `EMOJI_REGEX` is global (`/g`), so `lastIndex` is reset before matching
- * to ensure deterministic behavior across repeated calls.
- *
- * @param {string} text - Input string potentially containing emoji sequences.
- * @returns {string} A string containing only the matched emoji sequences,
- *                   concatenated in original order.
- */
-function extractEmojiSequences(text) {
-    EMOJI_REGEX.lastIndex = 0;
-    if (!EMOJI_REGEX.test(text)) return ''; // quick reject
-    EMOJI_REGEX.lastIndex = 0;
-
-    let out = '';
-    for (const match of text.matchAll(EMOJI_REGEX)) out += match[0];
-    return out;
-}
-
-/**
- * Matches any character that is NOT:
- * - Printable ASCII (U+0020–U+007E)
- * - Newline characters: \n (LF), \r (CR)
- * This allows for \n, \r, and \r\n line endings.
- */
-export const ASCII_KEYBOARD_SAFE_REGEX = /[^\x20-\x7E\r\n]+/gu;
-
-/**
- * Removes all non-keyboard characters, preserving:
- * - Printable ASCII (0x20–0x7E)
- * - Emojis (Unicode Emoji_Presentation + FE0F variants)
- */
-function purgeNonKeyboardChars(text, purgeEmojis = false) {
-    const normalized = normalizeTypographicJank(text);
-
-    // If emojis are being purged, keyboard-only becomes "ASCII + CR/LF only"
-    if (purgeEmojis) {
-        return normalized.replace(ASCII_KEYBOARD_SAFE_REGEX, '');
-    }
-
-    // Replace each non-ASCII run with ONLY the emoji sequences inside it.
-    // This preserves ZWJ sequences and VS16 variants (e.g. 👨‍👩‍👧‍👦, ✌️, ‼️),
-    // while dropping non-emoji non-ASCII (e.g. 𝌆).
-    return normalized.replace(ASCII_KEYBOARD_SAFE_REGEX, (run) => extractEmojiSequences(run));
-}
-
-
-// Smart apostrophes (single quote variations)
-const SMART_SINGLE_QUOTES_REGEX = /[\u2018\u2019\u201A\u201B\u2032\u2035]/g;
-
-// Smart double quotes (including primes and guillemets)
-const SMART_DOUBLE_QUOTES_REGEX = /[\u201C\u201D\u201E\u201F\u2033\u2036\u00AB\u00BB]/g;
-
-// Dash variants: en dash, em dash, horizontal bar, figure dash, minus
-const UNICODE_DASHES_REGEX = /[\u2012\u2013\u2014\u2015\u2212]/g;
-
-// Ellipsis character
-const ELLIPSIS_REGEX = /\u2026/g;
-
-// Bullets and middle dot
-const BULLETS_REGEX = /[\u2022\u00B7]/g;
-
-// Full-width ASCII punctuation: U+FF01 - U+FF5E
-const FULLWIDTH_PUNCTUATION_REGEX = /[\uFF01-\uFF5E]/g;
-
-/**
- * Normalizes typographic Unicode punctuation into ASCII equivalents.
- * @param {string} text
- * @returns {string}
- */
-export function normalizeTypographicJank(text) {
-    return text
-        .replace(SMART_SINGLE_QUOTES_REGEX, "'")
-        .replace(SMART_DOUBLE_QUOTES_REGEX, '"')
-        .replace(UNICODE_DASHES_REGEX, '-')
-        .replace(ELLIPSIS_REGEX, '...')
-        .replace(BULLETS_REGEX, '*')
-        .replace(FULLWIDTH_PUNCTUATION_REGEX, m =>
-            String.fromCharCode(m.charCodeAt(0) - 0xFEE0)
-        );
-}
-
-
-/** @type {RegExp} */
-export let EMOJI_REGEX;
-
-/**
- * Try Unicode property escape regex (preferred).
- * Fallback to basic emoji ranges if unsupported.
- */
-try {
-    EMOJI_REGEX = new RegExp(
-        '(?:\\p{Extended_Pictographic}(?:\\uFE0F|\\uFE0E)?(?:\\u200D(?:\\p{Extended_Pictographic}|\\w)+)*)',
-        'gu'
-    );
-} catch {
-    // Fallback: wide-range emoji component match (flags, tones, symbols)
-    //   * Covers:
-    //   * - Emoji base chars (1F300–1FAFF)
-    //   * - Dingbats (2700–27BF) like ❌, ✅, ☑️
-    //   * - Skin tones (1F3FB–1F3FF)
-    //   * - ZWJ (200D), variation selectors (FE0F)
-    //   * - Regional indicators (1F1E6–1F1FF)
-    EMOJI_REGEX = /[\u{2700}-\u{27BF}\u{1F300}-\u{1FAFF}\u{1F3FB}-\u{1F3FF}\u200D\uFE0F\u{1F1E6}-\u{1F1FF}]/gu;
-}
-
-/**
- * Removes all emoji characters using Unicode property escapes.
- * Supports modern environments (Unicode v13+) with fallback.
- *
- * @param {string} text
- * @returns {string}
- */
-function purgeEmojiCharacters(text) {
-    return text.replace(EMOJI_REGEX, '');
-}
-
-
-/**
- * Normalizes all line endings to a consistent format.
- * 
- * Converts:
- * - Windows ("\r\n"), Old Mac ("\r"), Unix ("\n")
- * Into the specified newline format (default: Unix "\n").
- *
- * @param {string} text - Input string to normalize.
- * @param {string} [normalized='\n'] - Target newline style (e.g. '\n', '\r\n').
- * @returns {string}
- */
-const NORMALIZE_NEWLINES_REGEX = /\r\n|\r|\n/g;
-function normalizeNewlineChars(text, normalized = '\n') {
-    return text.replace(NORMALIZE_NEWLINES_REGEX, normalized);
-}
-
-
-/**
- * Trims spaces and tabs immediately before and after newlines.
- * 
- * Example: "hello   \n   world" → "hello\nworld"
- *
- * Preserves the newline itself, just removes whitespace around it.
- *
- * @param {string} text
- * @returns {string}
- */
-const TRIM_SPACES_AROUND_NEWLINES_REGEX = /[ \t]*(\n+)[ \t]*/g;
-function trimSpacesAroundNewlineChars(text) {
-    return text.replace(TRIM_SPACES_AROUND_NEWLINES_REGEX, '$1');
-}
-
-
-/**
- * Collapses excessive newlines based on paragraph preservation settings.
- *
- * - If preserveParagraphs = true:
- *     Collapses 3 or more consecutive newlines → exactly two ("\n\n")
- * - If preserveParagraphs = false:
- *     Collapses 2 or more consecutive newlines → exactly one ("\n")
- * 
- * Example (preserveParagraphs = true):
- * "Line1\n\n\n\nLine2" → "Line1\n\nLine2"
- *
- * @param {string} text
- * @param {boolean} preserveParagraphs - Whether to preserve paragraph breaks (double newlines).
- * @returns {string}
- */
-const MULTIPLE_NEWLINES_REGEX = /\n{2,}/g;
-const TRIPLE_NEWLINES_REGEX = /\n{3,}/g;
-
-function collapseMultipleNewLines(text, preserveParagraphs) {
-    return preserveParagraphs
-        ? text.replace(TRIPLE_NEWLINES_REGEX, '\n\n')
-        : text.replace(MULTIPLE_NEWLINES_REGEX, '\n');
-}
-
-
-/**
- * Collapses multiple consecutive spaces into a single space.
- * (Does not touch tabs or newlines.)
- * 
- * Example: "hello   world" → "hello world"
- *
- * @param {string} text
- * @returns {string}
- */
-const MULTIPLE_SPACES_REGEX = / {2,}/g;
-function collapseExtraSpaces(text) {
-    return text.replace(MULTIPLE_SPACES_REGEX, ' ');
-}
-
-
-/**
- * Nukes hidden control characters that are invisible and often dangerous.
- * (Excludes necessary whitespace like \n and \t.)
- * 
- * Control characters nuked (the code points come from the table in scripts/categories.js):
- *   - NUL and the C0 controls except tab, LF and CR; DEL (U+007F)
- *   - C1 controls (U+0080-U+009F)
- *   - Bidi controls
- * 
- * @param {string} text
- * @returns {string}
- */
-function purgeControlCharacters(text) {
-    return text.replace(CONTROL_CHARS_REGEX, '');
-}
+// src/sanctifyText.js
+
+import { CONTROL_CHARS_REGEX, INVISIBLE_TRASH_REGEX } from './categories.generated.js';
+export { CONTROL_CHARS_REGEX, INVISIBLE_TRASH_REGEX };
+
+
+/**
+ * @typedef {Object} SanctifyOptions
+ * @property {boolean} [purgeInvisibleChars]
+ * @property {boolean} [purgeEmojis]
+ * @property {boolean} [nukeControls]
+ * @property {boolean} [keyboardOnlyFilter]
+ * @property {boolean} [normalizeNewlines]
+ * @property {boolean} [trimSpacesAroundNewlines]
+ * @property {boolean} [collapseNewLines]
+ * @property {boolean} [preserveParagraphs]
+ * @property {boolean} [collapseSpaces]
+ * @property {boolean} [finalTrim]
+ * @property {number} [maxLength]
+ * @property {'throw'|'truncate'|'noop'} [onMaxLength]
+ * @property {boolean} [normalizeSpaces] - Replace look-alike spaces with an ordinary space (runs first).
+ * @property {boolean} [purgeZeroWidthSpaces] - Remove U+200B, U+2060 and U+FEFF only, keeping joiners and selectors.
+ */
+
+
+/**
+ * Summons a customized sanctifier function with pre-bound booleans.
+ *
+ * Accepts full flag names and returns a text-cleaning function.
+ *
+ * @param {Object} [defaultOptions={}]
+ * @param {boolean} [defaultOptions.purgeInvisibleChars]
+ * @param {boolean} [defaultOptions.purgeEmojis]
+ * @param {boolean} [defaultOptions.nukeControls]
+ * @param {boolean} [defaultOptions.keyboardOnlyFilter]
+ * @param {boolean} [defaultOptions.normalizeNewlines]
+ * @param {boolean} [defaultOptions.trimSpacesAroundNewlines]
+ * @param {boolean} [defaultOptions.collapseNewLines]
+ * @param {boolean} [defaultOptions.preserveParagraphs]
+ * @param {boolean} [defaultOptions.collapseSpaces]
+ * @param {boolean} [defaultOptions.finalTrim]
+ * @param {number} [defaultOptions.maxLength=Infinity] - Hard input length cap (UTF-16 code units).
+ * @param {'throw' | 'truncate' | 'noop'} [defaultOptions.onMaxLength='throw'] - Behavior when input exceeds maxLength.
+ * @param {boolean} [defaultOptions.normalizeSpaces]
+ * @param {boolean} [defaultOptions.purgeZeroWidthSpaces]
+ * @returns {(text: string) => string}
+ */
+export function summonSanctifier(defaultOptions = {}) {
+    const normalizeSpaces = !!defaultOptions.normalizeSpaces;
+    const purgeZeroWidthSpaces = !!defaultOptions.purgeZeroWidthSpaces;
+    const purgeInvisibleChars = !!defaultOptions.purgeInvisibleChars;
+    const purgeEmojis = !!defaultOptions.purgeEmojis;
+    const nukeControls = !!defaultOptions.nukeControls;
+    const keyboardOnlyFilter = !!defaultOptions.keyboardOnlyFilter;
+    const normalizeNewlines = !!defaultOptions.normalizeNewlines;
+    const trimSpacesAroundNewlines = !!defaultOptions.trimSpacesAroundNewlines;
+    const collapseNewLines = !!defaultOptions.collapseNewLines;
+    const preserveParagraphs = !!defaultOptions.preserveParagraphs;
+    const collapseSpaces = !!defaultOptions.collapseSpaces;
+    const finalTrim = !!defaultOptions.finalTrim;
+
+    const maxLength =
+        Number.isFinite(defaultOptions.maxLength) && defaultOptions.maxLength >= 0
+            ? defaultOptions.maxLength
+            : Infinity;
+
+    const onMaxLength =
+        defaultOptions.onMaxLength === 'truncate' ||
+            defaultOptions.onMaxLength === 'noop' ||
+            defaultOptions.onMaxLength === 'throw'
+            ? defaultOptions.onMaxLength
+            : 'throw';
+
+    return (text) =>
+        sanctifyText(
+            text,
+            purgeInvisibleChars,
+            purgeEmojis,
+            nukeControls,
+            keyboardOnlyFilter,
+            normalizeNewlines,
+            trimSpacesAroundNewlines,
+            collapseNewLines,
+            preserveParagraphs,
+            collapseSpaces,
+            finalTrim,
+            maxLength,
+            onMaxLength,
+            normalizeSpaces,
+            purgeZeroWidthSpaces
+        );
+}
+
+// --- Added Presets ---
+
+/**
+ * Strict sanitizer:
+ * - Purge emojis
+ * - Collapse all newlines
+ * - Collapse spaces
+ * - Nuke control characters
+ */
+/** @param {string} text @returns {string} */
+summonSanctifier.strict = text => sanctifyText(
+    text,
+    true,   // purgeInvisibleChars
+    true,   // purgeEmojis
+    true,   // nukeControls
+    false,  // keyboardOnlyFilter
+    true,   // normalizeNewlines
+    true,   // trimSpacesAroundNewlines
+    true,   // collapseNewLines
+    false,  // preserveParagraphs
+    true,   // collapseSpaces
+    true    // finalTrim
+);
+
+
+/**
+ * Loose sanitizer:
+ * - Collapse spaces
+ * - Preserve paragraphs
+ * - Normalize newlines
+ */
+/** @param {string} text @returns {string} */
+summonSanctifier.loose = text => sanctifyText(
+    text,
+    false,  // purgeInvisibleChars
+    false,  // purgeEmojis
+    false,  // nukeControls
+    false,  // keyboardOnlyFilter
+    true,   // normalizeNewlines
+    true,   // trimSpacesAroundNewlines
+    true,   // collapseNewLines
+    true,   // preserveParagraphs
+    true,   // collapseSpaces
+    true    // finalTrim
+);
+
+
+/**
+ * Keyboard-only (with emojis):
+ * - Keeps emojis and printable ASCII
+ * - Strips non-standard characters
+ * - Normalizes typographic trash
+ */
+/** @param {string} text @returns {string} */
+summonSanctifier.keyboardOnlyEmoji = text => sanctifyText(
+    text,
+    false,  // purgeInvisibleChars
+    false,  // purgeEmojis
+    false,  // nukeControls
+    true,   // keyboardOnlyFilter
+    true,   // normalizeNewlines
+    true,   // trimSpacesAroundNewlines
+    false,  // collapseNewLines
+    false,  // preserveParagraphs
+    false,  // collapseSpaces
+    true    // finalTrim
+);
+
+
+/**
+ * Keyboard-only (strict):
+ * - Removes emojis
+ * - Collapses all whitespace
+ * - Restricts to printable ASCII only
+ */
+/** @param {string} text @returns {string} */
+summonSanctifier.keyboardOnly = text => sanctifyText(
+    text,
+    true,   // purgeInvisibleChars
+    true,   // purgeEmojis
+    true,   // nukeControls
+    true,   // keyboardOnlyFilter
+    true,   // normalizeNewlines
+    true,   // trimSpacesAroundNewlines
+    true,   // collapseNewLines
+    false,  // preserveParagraphs
+    true,   // collapseSpaces
+    true    // finalTrim
+);
+
+
+/**
+ * Pasted text and model output:
+ * - Look-alike spaces (NBSP, narrow NBSP, typographic spaces) become ordinary spaces,
+ *   so words never run together
+ * - Zero-width spaces, word joiners and BOMs are removed
+ * - Newlines are normalized to `\n`
+ * Keeps emoji joiners, variation selectors, bidi marks, indentation and line breaks.
+ */
+/** @param {string} text @returns {string} */
+summonSanctifier.pasted = text => sanctifyText(
+    text,
+    false,  // purgeInvisibleChars
+    false,  // purgeEmojis
+    false,  // nukeControls
+    false,  // keyboardOnlyFilter
+    true,   // normalizeNewlines
+    false,  // trimSpacesAroundNewlines
+    false,  // collapseNewLines
+    false,  // preserveParagraphs
+    false,  // collapseSpaces
+    false,  // finalTrim
+    Infinity, // maxLength
+    'throw',  // onMaxLength
+    true,   // normalizeSpaces
+    true    // purgeZeroWidthSpaces
+);
+
+
+/**
+ * Text Sanctifier
+ *
+ * Brutal text normalizer and invisible trash scrubber,
+ * configurable to kill whatever ghosts you want dead.
+ *
+ * ⚠️ Note: This is plain-text normalization/filtering — not an HTML/XSS sanitizer.
+ *
+ * @param {string} text
+ * @param {boolean} [purgeInvisibleChars=false] - Remove ZWSP, NBSP, bidi, tags, variation selectors, etc.
+ * @param {boolean} [purgeEmojis=false] - Remove emoji characters entirely.
+ * @param {boolean} [nukeControls=false] - Remove non-whitespace control characters.
+ * @param {boolean} [keyboardOnlyFilter=false] - Keep printable ASCII + full emoji sequences only (drops other Unicode).
+ * @param {boolean} [normalizeNewlines=false] - Convert all newlines to `\n`.
+ * @param {boolean} [trimSpacesAroundNewlines=false] - Remove spaces/tabs around newlines.
+ * @param {boolean} [collapseNewLines=false] - Collapse `\n` runs (optionally preserve paragraphs).
+ * @param {boolean} [preserveParagraphs=false] - Preserve paragraph breaks when collapsing newlines.
+ * @param {boolean} [collapseSpaces=false] - Collapse multiple spaces into one.
+ * @param {boolean} [finalTrim=false] - `.trim()` the final output (head/tail).
+ * @param {number} [maxLength=Infinity] - Hard input length cap (UTF-16 code units via `text.length`).
+ * @param {'throw'|'truncate'|'noop'} [onMaxLength='throw'] - Behavior when `text.length` exceeds `maxLength`:
+ *   - `'throw'`: throw a `RangeError`
+ *   - `'truncate'`: slice to `maxLength` before processing
+ *   - `'noop'`: return the original input unchanged (skips processing)
+ * @param {boolean} [normalizeSpaces=false] - Replace look-alike spaces with an ordinary space. Runs first,
+ *   so `purgeInvisibleChars` no longer joins the words around them.
+ * @param {boolean} [purgeZeroWidthSpaces=false] - Remove zero-width spaces, word joiners and BOMs only.
+ * @returns {string}
+ * @throws {TypeError} If `text` is not a string.
+ * @throws {RangeError} If input exceeds `maxLength` and `onMaxLength` is `'throw'`.
+ */
+export function sanctifyText(
+    text,
+    purgeInvisibleChars = false,
+    purgeEmojis = false,
+    nukeControls = false,
+    keyboardOnlyFilter = false,
+    normalizeNewlines = false,
+    trimSpacesAroundNewlines = false,
+    collapseNewLines = false,
+    preserveParagraphs = false,
+    collapseSpaces = false,
+    finalTrim = false,
+    maxLength = Infinity,
+    onMaxLength = 'throw',
+    normalizeSpaces = false,
+    purgeZeroWidthSpaces = false
+) {
+    if (typeof text !== 'string') {
+        throw new TypeError('sanctifyText expects a string input.');
+    }
+
+    if (onMaxLength !== 'truncate' && onMaxLength !== 'noop' && onMaxLength !== 'throw') {
+        onMaxLength = 'throw';
+    }
+
+    if (text.length > maxLength) {
+        switch (onMaxLength) {
+            case 'truncate':
+                text = text.slice(0, maxLength);
+                break;
+            case 'noop':
+                return text;
+            case 'throw':
+            default:
+                throw new RangeError(
+                    `sanctifyText input length ${text.length} exceeds maxLength ${maxLength}.`
+                );
+        }
+    }
+
+    let cleaned = text;
+
+    // Look-alike spaces become ordinary spaces, before anything can delete them
+    if (normalizeSpaces) cleaned = cleaned.replace(SPACE_LOOKALIKES_REGEX, ' ');
+
+    // Zero-width spaces, word joiners and BOMs (not joiners or selectors)
+    if (purgeZeroWidthSpaces) cleaned = cleaned.replace(ZERO_WIDTH_SPACES_REGEX, '');
+
+    //  Purge invisible Unicode trash (zero-width, non-breaking, bidi junk, etc.)
+    if (purgeInvisibleChars) cleaned = purgeInvisibleTrash(cleaned);
+
+    // Remove emojis
+    if (purgeEmojis) cleaned = purgeEmojiCharacters(cleaned);
+
+    // Nuke control characters (excluding whitespace)
+    if (nukeControls) cleaned = purgeControlCharacters(cleaned);
+
+    // Keep only ASCII/emojis
+    if (keyboardOnlyFilter) cleaned = purgeNonKeyboardChars(cleaned, purgeEmojis);
+
+    // Normalize line endings to Unix style (\n)
+    if (normalizeNewlines) cleaned = normalizeNewlineChars(cleaned);
+
+    // Remove spaces/tabs around newlines
+    if (trimSpacesAroundNewlines) cleaned = trimSpacesAroundNewlineChars(cleaned);
+
+    // Collapse excessive newlines, Optionally preserve Paragraphs
+    if (collapseNewLines) cleaned = collapseMultipleNewLines(cleaned, preserveParagraphs);
+
+    // Collapse multiple spaces into a single space
+    if (collapseSpaces) cleaned = collapseExtraSpaces(cleaned);
+
+    // Final trim, return Sanctified Text
+    return finalTrim ? cleaned.trim() : cleaned;
+}
+
+
+// --- Micro helpers ---
+
+
+/**
+ * Look-alike spaces: the `nonbreaking_space` category except U+3000, the ordinary
+ * space of CJK text. A test checks this against the category table.
+ */
+export const SPACE_LOOKALIKES_REGEX = /[\u00A0\u2000-\u200A\u202F\u205F]/g;
+
+/** Zero-width space, word joiner and BOM: invisible and never needed in plain text. */
+export const ZERO_WIDTH_SPACES_REGEX = /[\u200B\u2060\uFEFF]/g;
+
+
+/**
+ * Removes the invisible Unicode characters in INVISIBLE_TRASH_REGEX: non-breaking and
+ * typographic spaces, zero-width characters, bidi controls, invisible format characters,
+ * Hangul fillers, variation selectors (including VS16 in emoji), the BOM and tag characters.
+ * The code points come from the table in scripts/categories.js.
+ *
+ * @param {string} text
+ * @returns {string}
+ */
+function purgeInvisibleTrash(text) {
+    return text.replace(INVISIBLE_TRASH_REGEX, '');
+}
+
+
+/**
+ * Extracts and preserves complete emoji sequences from a string.
+ *
+ * Uses the global `EMOJI_REGEX` to match full emoji grapheme sequences,
+ * including:
+ *   - Extended pictographic characters
+ *   - Zero-width joiner (ZWJ) sequences (e.g. 👨‍👩‍👧‍👦)
+ *   - Variation Selector-16 (VS16) emoji presentation forms (e.g. ✌️, ‼️)
+ *
+ * Any non-emoji characters are ignored.
+ *
+ * Important:
+ * `EMOJI_REGEX` is global (`/g`), so `lastIndex` is reset before matching
+ * to ensure deterministic behavior across repeated calls.
+ *
+ * @param {string} text - Input string potentially containing emoji sequences.
+ * @returns {string} A string containing only the matched emoji sequences,
+ *                   concatenated in original order.
+ */
+function extractEmojiSequences(text) {
+    EMOJI_REGEX.lastIndex = 0;
+    if (!EMOJI_REGEX.test(text)) return ''; // quick reject
+    EMOJI_REGEX.lastIndex = 0;
+
+    let out = '';
+    for (const match of text.matchAll(EMOJI_REGEX)) out += match[0];
+    return out;
+}
+
+/**
+ * Matches any character that is NOT:
+ * - Printable ASCII (U+0020–U+007E)
+ * - Newline characters: \n (LF), \r (CR)
+ * This allows for \n, \r, and \r\n line endings.
+ */
+export const ASCII_KEYBOARD_SAFE_REGEX = /[^\x20-\x7E\r\n]+/gu;
+
+/**
+ * Removes all non-keyboard characters, preserving:
+ * - Printable ASCII (0x20–0x7E)
+ * - Emojis (Unicode Emoji_Presentation + FE0F variants)
+ */
+function purgeNonKeyboardChars(text, purgeEmojis = false) {
+    const normalized = normalizeTypographicJank(text);
+
+    // If emojis are being purged, keyboard-only becomes "ASCII + CR/LF only"
+    if (purgeEmojis) {
+        return normalized.replace(ASCII_KEYBOARD_SAFE_REGEX, '');
+    }
+
+    // Replace each non-ASCII run with ONLY the emoji sequences inside it.
+    // This preserves ZWJ sequences and VS16 variants (e.g. 👨‍👩‍👧‍👦, ✌️, ‼️),
+    // while dropping non-emoji non-ASCII (e.g. 𝌆).
+    return normalized.replace(ASCII_KEYBOARD_SAFE_REGEX, (run) => extractEmojiSequences(run));
+}
+
+
+// Smart apostrophes (single quote variations)
+const SMART_SINGLE_QUOTES_REGEX = /[\u2018\u2019\u201A\u201B\u2032\u2035]/g;
+
+// Smart double quotes (including primes and guillemets)
+const SMART_DOUBLE_QUOTES_REGEX = /[\u201C\u201D\u201E\u201F\u2033\u2036\u00AB\u00BB]/g;
+
+// Dash variants: en dash, em dash, horizontal bar, figure dash, minus
+const UNICODE_DASHES_REGEX = /[\u2012\u2013\u2014\u2015\u2212]/g;
+
+// Ellipsis character
+const ELLIPSIS_REGEX = /\u2026/g;
+
+// Bullets and middle dot
+const BULLETS_REGEX = /[\u2022\u00B7]/g;
+
+// Full-width ASCII punctuation: U+FF01 - U+FF5E
+const FULLWIDTH_PUNCTUATION_REGEX = /[\uFF01-\uFF5E]/g;
+
+/**
+ * Normalizes typographic Unicode punctuation into ASCII equivalents.
+ * @param {string} text
+ * @returns {string}
+ */
+export function normalizeTypographicJank(text) {
+    return text
+        .replace(SMART_SINGLE_QUOTES_REGEX, "'")
+        .replace(SMART_DOUBLE_QUOTES_REGEX, '"')
+        .replace(UNICODE_DASHES_REGEX, '-')
+        .replace(ELLIPSIS_REGEX, '...')
+        .replace(BULLETS_REGEX, '*')
+        .replace(FULLWIDTH_PUNCTUATION_REGEX, m =>
+            String.fromCharCode(m.charCodeAt(0) - 0xFEE0)
+        );
+}
+
+
+/** @type {RegExp} */
+export let EMOJI_REGEX;
+
+/**
+ * Try Unicode property escape regex (preferred).
+ * Fallback to basic emoji ranges if unsupported.
+ */
+try {
+    EMOJI_REGEX = new RegExp(
+        '(?:\\p{Extended_Pictographic}(?:\\uFE0F|\\uFE0E)?(?:\\u200D(?:\\p{Extended_Pictographic}|\\w)+)*)',
+        'gu'
+    );
+} catch {
+    // Fallback: wide-range emoji component match (flags, tones, symbols)
+    //   * Covers:
+    //   * - Emoji base chars (1F300–1FAFF)
+    //   * - Dingbats (2700–27BF) like ❌, ✅, ☑️
+    //   * - Skin tones (1F3FB–1F3FF)
+    //   * - ZWJ (200D), variation selectors (FE0F)
+    //   * - Regional indicators (1F1E6–1F1FF)
+    EMOJI_REGEX = /[\u{2700}-\u{27BF}\u{1F300}-\u{1FAFF}\u{1F3FB}-\u{1F3FF}\u200D\uFE0F\u{1F1E6}-\u{1F1FF}]/gu;
+}
+
+/**
+ * Removes all emoji characters using Unicode property escapes.
+ * Supports modern environments (Unicode v13+) with fallback.
+ *
+ * @param {string} text
+ * @returns {string}
+ */
+function purgeEmojiCharacters(text) {
+    return text.replace(EMOJI_REGEX, '');
+}
+
+
+/**
+ * Normalizes all line endings to a consistent format.
+ * 
+ * Converts:
+ * - Windows ("\r\n"), Old Mac ("\r"), Unix ("\n")
+ * Into the specified newline format (default: Unix "\n").
+ *
+ * @param {string} text - Input string to normalize.
+ * @param {string} [normalized='\n'] - Target newline style (e.g. '\n', '\r\n').
+ * @returns {string}
+ */
+const NORMALIZE_NEWLINES_REGEX = /\r\n|\r|\n/g;
+function normalizeNewlineChars(text, normalized = '\n') {
+    return text.replace(NORMALIZE_NEWLINES_REGEX, normalized);
+}
+
+
+/**
+ * Trims spaces and tabs immediately before and after newlines.
+ * 
+ * Example: "hello   \n   world" → "hello\nworld"
+ *
+ * Preserves the newline itself, just removes whitespace around it.
+ *
+ * @param {string} text
+ * @returns {string}
+ */
+const TRIM_SPACES_AROUND_NEWLINES_REGEX = /[ \t]*(\n+)[ \t]*/g;
+function trimSpacesAroundNewlineChars(text) {
+    return text.replace(TRIM_SPACES_AROUND_NEWLINES_REGEX, '$1');
+}
+
+
+/**
+ * Collapses excessive newlines based on paragraph preservation settings.
+ *
+ * - If preserveParagraphs = true:
+ *     Collapses 3 or more consecutive newlines → exactly two ("\n\n")
+ * - If preserveParagraphs = false:
+ *     Collapses 2 or more consecutive newlines → exactly one ("\n")
+ * 
+ * Example (preserveParagraphs = true):
+ * "Line1\n\n\n\nLine2" → "Line1\n\nLine2"
+ *
+ * @param {string} text
+ * @param {boolean} preserveParagraphs - Whether to preserve paragraph breaks (double newlines).
+ * @returns {string}
+ */
+const MULTIPLE_NEWLINES_REGEX = /\n{2,}/g;
+const TRIPLE_NEWLINES_REGEX = /\n{3,}/g;
+
+function collapseMultipleNewLines(text, preserveParagraphs) {
+    return preserveParagraphs
+        ? text.replace(TRIPLE_NEWLINES_REGEX, '\n\n')
+        : text.replace(MULTIPLE_NEWLINES_REGEX, '\n');
+}
+
+
+/**
+ * Collapses multiple consecutive spaces into a single space.
+ * (Does not touch tabs or newlines.)
+ * 
+ * Example: "hello   world" → "hello world"
+ *
+ * @param {string} text
+ * @returns {string}
+ */
+const MULTIPLE_SPACES_REGEX = / {2,}/g;
+function collapseExtraSpaces(text) {
+    return text.replace(MULTIPLE_SPACES_REGEX, ' ');
+}
+
+
+/**
+ * Nukes hidden control characters that are invisible and often dangerous.
+ * (Excludes necessary whitespace like \n and \t.)
+ * 
+ * Control characters nuked (the code points come from the table in scripts/categories.js):
+ *   - NUL and the C0 controls except tab, LF and CR; DEL (U+007F)
+ *   - C1 controls (U+0080-U+009F)
+ *   - Bidi controls
+ * 
+ * @param {string} text
+ * @returns {string}
+ */
+function purgeControlCharacters(text) {
+    return text.replace(CONTROL_CHARS_REGEX, '');
+}