text-sanctifier 1.3.1 → 1.3.2

package/README.md

@@ -7,7 +7,7 @@
 [![stars](https://img.shields.io/github/stars/iWhatty/text-sanctifier?style=social)](https://github.com/iWhatty/text-sanctifier)
 [![types](https://img.shields.io/npm/types/text-sanctifier)](https://www.npmjs.com/package/text-sanctifier)
 
-Brutal text normalizer and invisible Unicode scrubber for modern web projects. ~2.44 KB gzipped.
+Brutal text normalizer and invisible Unicode scrubber for modern web projects. ~2.57 KB gzipped.
 
 ## Features
 
@@ -15,7 +15,8 @@ Brutal text normalizer and invisible Unicode scrubber for modern web projects. ~
 - Purges bidi controls, including the isolates (U+2066–U+2069) and the Arabic letter mark (U+061C) used in Trojan Source attacks
 - Purges Unicode tag characters (U+E0000–U+E007F), the "ASCII smuggling" channel for hiding instructions from humans in text an LLM reads, plus variation selectors, Hangul fillers, soft hyphens and invisible operators
 - Locates invisible and control characters by offset, line and column (`inspectTextDetailed`)
-- `text-sanctifier check` / `fix` command line for CI and coding agents, with JSON output and exit codes
+- `text-sanctifier check` / `fix` command line for CI and coding agents, with JSON or SARIF output, exit codes and pre-commit hooks
+- `check --ai-config` finds and decodes hidden text in agent instruction files (`CLAUDE.md`, `AGENTS.md`, Cursor rules, `SKILL.md`, MCP configuration)
 - Normalizes line endings (CRLF, CR, LF) to LF
 - Collapses unwanted spaces and paragraphs
 - Nukes control characters (if enabled)
@@ -163,23 +164,25 @@ truncated: false
 */
 ```
 
-| Category | Code points |
-| --- | --- |
-| `bidi_control` | U+061C, U+200E, U+200F, U+202A–U+202E, U+2066–U+2069 |
-| `zero_width` | U+200B–U+200D, U+2060 |
-| `nonbreaking_space` | U+00A0, U+2000–U+200A, U+202F, U+205F, U+3000 (non-breaking and typographic spaces) |
-| `bom` | U+FEFF; `atStart` is true at offset 0, where a BOM is legitimate |
-| `nul` | U+0000 |
-| `c0_control` | U+0001–U+001F except tab, LF and CR; U+007F |
-| `c1_control` | U+0080–U+009F |
-| `invisible_format` | U+00AD (soft hyphen), U+034F, U+180E, U+2061–U+2064 (invisible operators) |
-| `hangul_filler` | U+115F, U+1160, U+3164, U+FFA0 (letters that render blank) |
-| `variation_selector` | U+FE00–U+FE0F, U+E0100–U+E01EF |
-| `tag` | U+E0000–U+E007F (tag characters, used for "ASCII smuggling" of hidden instructions) |
+<!-- categories:start (generated from scripts/categories.js by `npm run gen`; do not edit) -->
+| Category | Code points | Removed by |
+| --- | --- | --- |
+| `bidi_control` | U+061C, U+200E, U+200F, U+202A–U+202E, U+2066–U+2069 | `nukeControls`, `purgeInvisibleChars` |
+| `zero_width` | U+200B–U+200D, U+2060 | `purgeInvisibleChars` |
+| `nonbreaking_space` | U+00A0, U+2000–U+200A, U+202F, U+205F, U+3000 (non-breaking and typographic spaces) | `purgeInvisibleChars` |
+| `bom` | U+FEFF (`atStart` is true at offset 0, where a BOM is legitimate) | `purgeInvisibleChars` |
+| `nul` | U+0000 | `nukeControls` |
+| `c0_control` | U+0001–U+0008, U+000B, U+000C, U+000E–U+001F, U+007F (tab, LF and CR are whitespace, not trash) | `nukeControls` |
+| `c1_control` | U+0080–U+009F | `nukeControls` |
+| `invisible_format` | U+00AD (soft hyphen), U+034F (combining grapheme joiner), U+17B4, U+17B5 (Khmer inherent vowels), U+180E (Mongolian vowel separator), U+2061–U+2064 (invisible operators), U+2065 (reserved), U+206A–U+206F (deprecated format characters), U+FFF0–U+FFF8 (reserved), U+1BCA0–U+1BCA3 (shorthand format controls), U+1D173–U+1D17A (musical formatting), U+E0080–U+E00FF (reserved), U+E01F0–U+E0FFF (reserved) (every other Default_Ignorable_Code_Point, which renderers draw as nothing) | `purgeInvisibleChars` |
+| `hangul_filler` | U+115F, U+1160, U+3164, U+FFA0 (letters that render blank) | `purgeInvisibleChars` |
+| `variation_selector` | U+180B–U+180D (Mongolian FVS1-FVS3), U+180F (Mongolian FVS4), U+FE00–U+FE0F (VS1-VS16, including VS16 in emoji), U+E0100–U+E01EF (VS17-VS256) | `purgeInvisibleChars` |
+| `tag` | U+E0000–U+E007F (tag characters, used for "ASCII smuggling" of hidden instructions) | `purgeInvisibleChars` |
+<!-- categories:end -->
 
 `codePoint` is the full code point, so a tag character reports `0xE0041`, not a surrogate. `offset` is a 0-based UTF-16 index (for a code point above U+FFFF, that of its high surrogate); `line` and `column` are 1-based, with columns in UTF-16 code units and lines ending at LF, CR or CRLF. Every occurrence is counted; only the first `maxFindings` (default 100, `Infinity` for all) are listed, and `truncated` says when some were not. The scan is one pass with no regex and no copy of the input.
 
-**Choosing `maxFindings`:** the scan itself uses almost no memory, but every listed finding is an object, so the cost of `maxFindings` grows with the number of findings, not the size of the input. 16 MB of non-breaking spaces is 16.7 million findings. With a cap of 20, it scans in about 70 ms and uses 0.2 MB. With `Infinity`, expect roughly 100–200 bytes per finding: about 1.5 GB for 8 million. On untrusted input, keep a cap. Use `maxFindings: 0` for a counts-only report: it builds no finding objects, and `counts`, `total` and `truncated` stay exact.
+**Choosing `maxFindings`:** the scan itself uses almost no memory, but every listed finding is an object, so the cost of `maxFindings` grows with the number of findings, not the size of the input. 16 MB of non-breaking spaces is 16.7 million findings. With a cap of 20, it scans in about 80 ms and uses 0.2 MB. With `Infinity`, expect roughly 100–200 bytes per finding: about 1.5 GB for 8 million. On untrusted input, keep a cap. Use `maxFindings: 0` for a counts-only report: it builds no finding objects, and `counts`, `total` and `truncated` stay exact.
 
 ### `getNewlineStats(text: string): NewlineStats`
 
@@ -191,9 +194,9 @@ import { getNewlineStats } from 'text-sanctifier';
 getNewlineStats('a\r\nb\n'); // { crlf: 1, cr: 0, lf: 1, types: ['CRLF', 'LF'], mixed: true }
 ```
 
-`zero_width` also flags U+200D (zero width joiner) inside emoji ZWJ sequences such as 👨‍👩‍👧, so a finding there is not necessarily hidden text. Filter by neighbouring code points if you need to tell them apart. (`purgeInvisibleChars` removes those joiners too, splitting the emoji; `keyboardOnlyEmoji` keeps the sequences whole.) The same holds for `variation_selector`, which flags VS16 (U+FE0F) in emoji such as ✌️ and the ideographic variation selectors in CJK names, and for `tag`, which flags the tag characters inside subdivision flags such as the England flag (U+1F3F4 followed by tags). `purgeInvisibleChars` strips them all; `keyboardOnlyEmoji` keeps VS16.
+`zero_width` also flags U+200D (zero width joiner) inside emoji ZWJ sequences such as 👨‍👩‍👧, so a finding there is not necessarily hidden text. Filter by neighbouring code points if you need to tell them apart. (`purgeInvisibleChars` removes those joiners too, splitting the emoji; `keyboardOnlyEmoji` keeps the sequences whole.) The same holds for `variation_selector`, which flags VS16 (U+FE0F) in emoji such as ✌️ and the ideographic variation selectors in CJK names, and for `tag`, which flags the tag characters inside subdivision flags such as the England flag (U+1F3F4 followed by tags). `purgeInvisibleChars` strips them all; `keyboardOnlyEmoji` keeps VS16. The Mongolian free variation selectors (U+180B–U+180D, U+180F) are also `variation_selector`: they are legitimate in Mongolian text, which `purgeInvisibleChars` can change, so don't purge Mongolian text with it.
 
-The categories cover exactly what `nukeControls` and `purgeInvisibleChars` strip.
+The categories cover exactly what `nukeControls` and `purgeInvisibleChars` strip. Every code point with the Unicode property Default_Ignorable_Code_Point (characters renderers must draw as nothing) is in a category, and a test checks this against the running Node's Unicode data.
 
 ---
 
@@ -216,9 +219,61 @@ src/auth.js:40:3  U+202E  bidi_control
 2 findings in 2 of 31 files.
 ```
 
-- `--json`: one JSON object on stdout, `{ files, total, filesScanned, filesWithFindings, skipped }`. Each entry in `files` is `{ path, ...inspectTextDetailed(text) }`; only files with findings are listed.
+- `--json` (or `--format json`): one JSON object on stdout, `{ files, total, filesScanned, filesWithFindings, skipped }`. Each entry in `files` is `{ path, ...inspectTextDetailed(text) }`; only files with findings are listed.
+- `--format sarif`: a [SARIF 2.1.0](https://docs.oasis-open.org/sarif/sarif/v2.1.0/sarif-v2.1.0.html) log for code-scanning viewers and editors. See [SARIF output](#sarif-output).
 - `--max-findings N`: findings listed per file (default 100; `all` for no cap). Counts always cover every occurrence.
 - `--strip-bom`: also report a BOM at offset 0, which is skipped by default.
+- `--ai-config`: check agent instruction files for hidden text. See [AI configuration files](#ai-configuration-files).
+
+### AI configuration files
+
+`check --ai-config` looks for hidden instructions in the files coding agents read: rules files, skill files and MCP configuration. Tag characters, variation selectors and invisible operators can carry text that people reviewing the file cannot see but a model reads.
+
+```sh
+npx text-sanctifier check --ai-config
+```
+
+```
+skills/review/SKILL.md:1:17  U+E0049  run of 15 tag characters spelling "Ignore the user"
+skills/review/SKILL.md:2:3  U+FE00  run of 3 variation selectors
+.cursor/rules/style.mdc:4:9  U+200B  zero_width
+19 findings (2 hidden runs) in 2 of 6 AI configuration files.
+```
+
+- With no paths it searches the current directory; directories you name are searched the same way, and files you name are checked whatever they are called. In a git work tree, files `.gitignore` excludes are skipped.
+- Files it looks for, matched case-insensitively: `CLAUDE.md`, `CLAUDE.local.md`, `AGENTS.md`, `AGENT.md`, `GEMINI.md` and `SKILL.md` at any depth; and at the top of the directory, `.claude/**`, `.cursorrules`, `.cursor/rules/**`, `.cursor/mcp.json`, `.windsurfrules`, `.windsurf/rules/**`, `.clinerules` (file or directory), `.roorules`, `.roo/rules/**`, `.github/copilot-instructions.md`, `.github/{instructions,prompts,chatmodes,agents}/**`, `.mcp.json`, `mcp.json` and `.vscode/mcp.json`.
+- Consecutive tag characters, variation selectors or invisible operators are reported once, as a run, at the first character. Tag runs are decoded: U+E0020–U+E007E become the ASCII they shadow, and other tags are shown as `<U+XXXX>`.
+- Every other invisible or control character is reported as in a plain `check`.
+- Two legitimate sequences are left out: the flags of England, Scotland and Wales (U+1F3F4, tag letters, U+E007F), and a single U+FE0E or U+FE0F after another character (text or emoji presentation). A plain `check` still reports both.
+- `--json` adds `hiddenRuns` (`{ kind, offset, line, column, codePoint, count, decoded?, decodedTruncated? }`, at most `--max-findings` per file) and `hiddenRunCount` to each file, and `hiddenRunCount` to the summary. `kind` is `tag`, `variation_selector` or `invisible_operator`; `decoded` is cut at 1000 characters. `total` and `counts` leave out the legitimate sequences.
+
+### SARIF output
+
+`check --format sarif` (with or without `--ai-config`) prints one SARIF 2.1.0 log with one result per listed finding:
+
+- `ruleId` is the category, and `tool.driver.rules` lists every category.
+- `level`: `error` for `bidi_control` and `tag`, `note` for `nonbreaking_space`, `warning` for the rest.
+- Each location has the file (relative to `%SRCROOT%`, the current directory), `startLine`, `startColumn` and `endColumn`. Columns count UTF-16 code units, as in `inspectTextDetailed` (`columnKind: "utf16CodeUnits"`).
+- With `--ai-config`, the message of the first result in a run describes the run and its decoded text.
+- `--max-findings` caps results per file as usual. `runs[0].properties` holds `total`, `filesScanned`, `filesWithFindings`, `truncated` (files with findings left out) and `skipped`.
+
+### pre-commit
+
+The repository publishes a [pre-commit](https://pre-commit.com) hook manifest, so the check runs locally on the files being committed:
+
+```yaml
+# .pre-commit-config.yaml
+repos:
+  - repo: https://github.com/iWhatty/text-sanctifier
+    rev: v1.3.2  # a release tag
+    hooks:
+      - id: text-sanctifier            # every staged text file
+      - id: text-sanctifier-ai-config  # agent instruction files, with --ai-config
+```
+
+- `text-sanctifier` runs `text-sanctifier check` on staged text files. Use `exclude:` for files that legitimately contain these characters, or `args: [--strip-bom]` and the other `check` options.
+- `text-sanctifier-ai-config` runs `text-sanctifier check --ai-config` on staged files that match the `--ai-config` file list.
+- A hook fails the commit when it finds something (exit code 1). Neither hook changes files; run `text-sanctifier fix` yourself.
 
 ### `fix [paths...]`
 
@@ -230,7 +285,7 @@ Rewrites files in place and reports what it removed.
 - With none of the three, `fix` uses `--purge-invisible --nuke-controls`, which removes exactly what `check` reports.
 - `--strip-bom`: also remove a BOM at offset 0. It is kept by default.
 - `--dry-run`: write nothing; report what would change.
-- `--json`: `{ files, dryRun, filesScanned, filesChanged, skipped }`, where each file is `{ path, removed, removedTotal, newlinesNormalized }`.
+- `--json` (or `--format json`): `{ files, dryRun, filesScanned, filesChanged, skipped }`, where each file is `{ path, removed, removedTotal, newlinesNormalized }`.
 
 `fix` deletes characters; it does not replace them. A non-breaking space between two words joins them, and emoji lose their zero width joiners and variation selectors. Run `check` or `fix --dry-run` first on prose.
 

package/bin/text-sanctifier.js

@@ -3,16 +3,19 @@
 //
 // Command-line front end for coding agents and CI:
 //   text-sanctifier check [paths...]   report invisible and control characters
+//   text-sanctifier check --ai-config  report hidden text in agent instruction files
 //   text-sanctifier fix [paths...]     remove them in place
 //
 // Plain Node (>=18), no dependencies. Exit codes: 0 clean, 1 findings (check,
 // or fix --dry-run with changes pending), 2 usage or I/O error.
 
 import { readFileSync, writeFileSync, statSync, readdirSync, existsSync } from 'node:fs';
-import { resolve, join, sep } from 'node:path';
+import { resolve, join, sep, relative } from 'node:path';
 import { spawnSync } from 'node:child_process';
+import { pathToFileURL } from 'node:url';
 import { inspectTextDetailed, summonSanctifier } from '../src/index.js';
 import { getNewlineStats } from '../src/inspectText.js';
+import { findHiddenRuns } from '../src/hiddenText.js';
 
 const VERSION = JSON.parse(readFileSync(new URL('../package.json', import.meta.url), 'utf8')).version;
 
@@ -27,9 +30,14 @@ alone). "-", or no paths with piped input, reads stdin. Directories and globs
 skip files ignored by git (inside a git work tree) and skip binary files.
 
 check options:
-  --json               JSON report on stdout
+  --format F           text (default), json or sarif (SARIF 2.1.0)
+  --json               same as --format json
   --max-findings N     findings listed per file (default 100, "all" for no cap)
   --strip-bom          also report a BOM at the start of a file
+  --ai-config          check agent instruction files (CLAUDE.md, AGENTS.md,
+                       .cursor/rules, SKILL.md, MCP config, ...) for hidden text;
+                       directories, or "." when no paths are given, are searched
+                       for those files, and tag text is decoded
 
 fix options:
   --purge-invisible    remove zero-width, non-breaking space, bidi and BOM characters
@@ -37,7 +45,8 @@ fix options:
   --normalize-newlines convert CRLF and CR to LF
   --strip-bom          also remove a BOM at the start of a file (kept by default)
   --dry-run            report what would change; write nothing
-  --json               JSON report
+  --format F           text (default) or json
+  --json               same as --format json
   With no cleaning flag, fix uses --purge-invisible --nuke-controls.
   stdin is cleaned to stdout; the report then goes to stderr.
 
@@ -62,7 +71,7 @@ class UsageError extends Error {}
 // --- Arguments ---
 
 const BOOLEAN_FLAGS = {
-    '--json': 'json',
+    '--ai-config': 'aiConfig',
     '--strip-bom': 'stripBom',
     '--purge-invisible': 'purgeInvisible',
     '--nuke-controls': 'nukeControls',
@@ -70,6 +79,8 @@ const BOOLEAN_FLAGS = {
     '--dry-run': 'dryRun',
 };
 const FIX_ONLY_FLAGS = new Set(['purgeInvisible', 'nukeControls', 'normalizeNewlines', 'dryRun']);
+const CHECK_ONLY_FLAGS = new Set(['aiConfig']);
+const FORMATS = new Set(['text', 'json', 'sarif']);
 
 /**
  * @param {string[]} argv
@@ -78,7 +89,8 @@ function parseArgs(argv) {
     const opts = {
         command: null,
         paths: [],
-        json: false,
+        format: null,
+        aiConfig: false,
         stripBom: false,
         purgeInvisible: false,
         nukeControls: false,
@@ -108,6 +120,15 @@ function parseArgs(argv) {
             opts.maxFindings = parseMaxFindings(value);
             continue;
         }
+        if (name === '--format' || (name === '--json' && inline === undefined)) {
+            const value = name === '--json' ? 'json' : inline !== undefined ? inline : argv[++i];
+            if (!FORMATS.has(value)) {
+                throw new UsageError(`--format expects text, json or sarif, got ${value === undefined ? 'nothing' : JSON.stringify(value)}.`);
+            }
+            if (opts.format !== null && opts.format !== value) throw new UsageError(`Conflicting formats: ${opts.format} and ${value}.`);
+            opts.format = value;
+            continue;
+        }
         if (name in BOOLEAN_FLAGS && inline === undefined) {
             opts[BOOLEAN_FLAGS[name]] = true;
             continue;
@@ -119,11 +140,12 @@ function parseArgs(argv) {
     if (opts.command !== 'check' && opts.command !== 'fix') {
         throw new UsageError(opts.command === null ? 'Missing command.' : `Unknown command: ${opts.command}`);
     }
-    if (opts.command === 'check') {
-        for (const [flag, key] of Object.entries(BOOLEAN_FLAGS)) {
-            if (FIX_ONLY_FLAGS.has(key) && opts[key]) throw new UsageError(`${flag} only applies to fix.`);
-        }
+    for (const [flag, key] of Object.entries(BOOLEAN_FLAGS)) {
+        if (opts.command === 'check' && FIX_ONLY_FLAGS.has(key) && opts[key]) throw new UsageError(`${flag} only applies to fix.`);
+        if (opts.command === 'fix' && CHECK_ONLY_FLAGS.has(key) && opts[key]) throw new UsageError(`${flag} only applies to check.`);
     }
+    if (opts.format === null) opts.format = 'text';
+    if (opts.command === 'fix' && opts.format === 'sarif') throw new UsageError('--format sarif only applies to check.');
     if (opts.command === 'fix' && !opts.purgeInvisible && !opts.nukeControls && !opts.normalizeNewlines) {
         opts.purgeInvisible = true;
         opts.nukeControls = true;
@@ -255,13 +277,37 @@ function walkGlob(cwd, glob) {
 }
 
 /**
- * Expands the path arguments into a de-duplicated input list.
+ * Agent instruction and configuration files, relative to the directory searched.
+ * Matched case-insensitively. Keep .pre-commit-hooks.yaml (the text-sanctifier-ai-config
+ * hook's `files`) in step; tests/cli.test.js checks both against the same paths.
+ */
+const AI_CONFIG_GLOBS = [
+    '**/CLAUDE.md', '**/CLAUDE.local.md', '**/AGENTS.md', '**/AGENT.md', '**/GEMINI.md', '**/SKILL.md',
+    '.claude/**',
+    '.cursorrules', '.cursor/rules/**', '.cursor/mcp.json',
+    '.windsurfrules', '.windsurf/rules/**',
+    '.clinerules', '.clinerules/**',
+    '.roorules', '.roo/rules/**',
+    '.github/copilot-instructions.md', '.github/instructions/**', '.github/prompts/**',
+    '.github/chatmodes/**', '.github/agents/**',
+    '.mcp.json', 'mcp.json', '.vscode/mcp.json',
+];
+
+const AI_CONFIG_RES = AI_CONFIG_GLOBS.map((g) => new RegExp(globToRegExp(g).source, 'i'));
+
+/** @param {string} rel - slash-separated path relative to the directory searched */
+const isAiConfigPath = (rel) => AI_CONFIG_RES.some((re) => re.test(rel));
+
+/**
+ * Expands the path arguments into a de-duplicated input list. With `aiConfig`,
+ * directories contribute only agent instruction files (AI_CONFIG_GLOBS).
  * @param {string[]} args
  * @param {string} cwd
  * @param {(msg: string) => void} warn
+ * @param {boolean} [aiConfig]
  * @returns {{ inputs: Array<{ path: string, explicit: boolean }>, errors: number }}
  */
-function collectInputs(args, cwd, warn) {
+function collectInputs(args, cwd, warn, aiConfig = false) {
     const inputs = [];
     const seen = new Set();
     let errors = 0;
@@ -291,11 +337,16 @@ function collectInputs(args, cwd, warn) {
         }
 
         if (inGit === undefined) inGit = isGitWorkTree(cwd);
-        const listed =
+        let listed =
             (inGit && gitListFiles(cwd, isGlob ? `:(glob)${toSlash(arg)}` : toSlash(arg))) ||
             (isGlob ? walkGlob(cwd, arg) : walk(cwd, arg));
 
-        if (listed.length === 0) warn(`${arg}: no files matched`);
+        if (aiConfig && !isGlob) {
+            const base = resolve(cwd, arg);
+            listed = listed.filter((f) => isAiConfigPath(toSlash(relative(base, resolve(cwd, f)))));
+        } else if (listed.length === 0) {
+            warn(`${arg}: no files matched`);
+        }
         for (const f of listed) add(f, false);
     }
     return { inputs, errors };
@@ -360,6 +411,161 @@ const hex = (cp) => 'U+' + cp.toString(16).toUpperCase().padStart(4, '0');
 const formatCounts = (counts) =>
     Object.entries(counts).filter(([, n]) => n > 0).map(([k, n]) => `${k} ${n}`).join(', ');
 
+/**
+ * @param {number} n
+ * @param {string} word
+ */
+const plural = (n, word) => `${n} ${word}${n === 1 ? '' : 's'}`;
+
+
+// --- check --ai-config ---
+
+const MAX_DECODED = 1000;
+const RUN_NOUNS = { tag: 'tag character', variation_selector: 'variation selector', invisible_operator: 'invisible operator' };
+
+/**
+ * inspect() for agent instruction files: drops findings that belong to legitimate
+ * sequences (subdivision flags, a lone emoji presentation selector) and adds the runs
+ * of tag characters, variation selectors and invisible operators, with tag text decoded.
+ * @param {string} text
+ * @param {number} maxFindings
+ * @param {boolean} stripBom
+ */
+function inspectAiConfig(text, maxFindings, stripBom) {
+    const full = inspect(text, Infinity, stripBom);
+    const { runs, benign } = findHiddenRuns(text);
+
+    const kept = [];
+    let b = 0;
+    for (const f of full.findings) {
+        while (b < benign.length && benign[b][1] <= f.offset) b++;
+        if (b < benign.length && benign[b][0] <= f.offset) { full.counts[f.category]--; continue; }
+        kept.push(f);
+    }
+
+    // Every run character is a finding, so each run starts at a kept finding
+    const hiddenRuns = [];
+    const loose = [];
+    let r = 0;
+    for (const f of kept) {
+        while (r < runs.length && runs[r].end <= f.offset) r++;
+        const run = runs[r];
+        if (!run || f.offset < run.offset) { loose.push(f); continue; }
+        if (f.offset !== run.offset) continue;
+        const entry = { kind: run.kind, offset: run.offset, line: f.line, column: f.column, codePoint: f.codePoint, count: run.count };
+        if (run.decoded !== undefined) {
+            entry.decoded = run.decoded.slice(0, MAX_DECODED);
+            entry.decodedTruncated = run.decoded.length > MAX_DECODED;
+        }
+        hiddenRuns.push(entry);
+    }
+
+    const findings = kept.slice(0, maxFindings);
+    return {
+        report: {
+            ...full,
+            findings,
+            total: kept.length,
+            truncated: kept.length > findings.length,
+            hiddenRuns: hiddenRuns.slice(0, maxFindings),
+            hiddenRunCount: hiddenRuns.length,
+        },
+        hiddenRuns,
+        loose,
+    };
+}
+
+/** @param {{ kind: string, count: number, decoded?: string, decodedTruncated?: boolean }} run */
+function describeRun(run) {
+    const what = plural(run.count, RUN_NOUNS[run.kind]);
+    if (run.decoded === undefined) return `run of ${what}`;
+    return `run of ${what} spelling ${JSON.stringify(run.decoded)}${run.decodedTruncated ? ' (cut short)' : ''}`;
+}
+
+
+// --- SARIF ---
+
+// Level and description per category, for SARIF rules. Every category needs an entry;
+// tests/cli.test.js checks the rules against inspectTextDetailed's counts.
+const CATEGORY_RULES = {
+    bidi_control: ['error', 'Bidirectional control character, which can make text display in a different order than it is read (Trojan Source)'],
+    zero_width: ['warning', 'Zero-width character'],
+    nonbreaking_space: ['note', 'Non-breaking or typographic space that looks like an ordinary space'],
+    bom: ['warning', 'Byte order mark (U+FEFF) that is not at the start of the file'],
+    nul: ['warning', 'NUL character'],
+    c0_control: ['warning', 'C0 control character other than tab, LF and CR'],
+    c1_control: ['warning', 'C1 control character'],
+    invisible_format: ['warning', 'Invisible format character, such as a soft hyphen or an invisible operator'],
+    hangul_filler: ['warning', 'Hangul filler, a letter that renders blank'],
+    variation_selector: ['warning', 'Variation selector, which can carry hidden data'],
+    tag: ['error', 'Tag character, which can hide ASCII text that people cannot see but language models read'],
+};
+
+const SARIF_SCHEMA = 'https://json.schemastore.org/sarif-2.1.0.json';
+const INFORMATION_URI = 'https://github.com/iWhatty/text-sanctifier';
+const ABSOLUTE_PATH = /^(?:[a-zA-Z]:)?[\\/]/;
+
+/** @param {string} label */
+function sarifArtifact(label) {
+    if (label === STDIN_LABEL) return { uri: 'stdin', description: { text: 'standard input' } };
+    if (ABSOLUTE_PATH.test(label)) return { uri: pathToFileURL(label).href };
+    return { uri: encodeURI(toSlash(label)).replace(/[?#]/g, encodeURIComponent), uriBaseId: '%SRCROOT%' };
+}
+
+/**
+ * Builds a SARIF 2.1.0 log with one result per listed finding; the rule is the category.
+ * @param {Array<{ path: string, findings: Array<{ offset: number, line: number, column: number, codePoint: number, category: string }>, hiddenRuns?: Array<{ offset: number, kind: string, count: number, decoded?: string }>, truncated: boolean }>} files
+ * @param {{ cwd: string, total: number, filesScanned: number, skipped: Array<{ path: string, reason: string }>, succeeded: boolean }} info
+ */
+function sarifLog(files, { cwd, total, filesScanned, skipped, succeeded }) {
+    const ruleIds = Object.keys(inspectTextDetailed('', { maxFindings: 0 }).counts);
+    const rules = ruleIds.map((id) => ({
+        id,
+        shortDescription: { text: CATEGORY_RULES[id][1] },
+        defaultConfiguration: { level: CATEGORY_RULES[id][0] },
+    }));
+
+    const results = [];
+    for (const f of files) {
+        const runAt = new Map((f.hiddenRuns || []).map((run) => [run.offset, run]));
+        const artifactLocation = sarifArtifact(f.path);
+        for (const x of f.findings) {
+            const run = runAt.get(x.offset);
+            results.push({
+                ruleId: x.category,
+                ruleIndex: ruleIds.indexOf(x.category),
+                level: CATEGORY_RULES[x.category][0],
+                message: { text: `${hex(x.codePoint)} ${x.category}${run ? `, start of a ${describeRun(run)}` : ''}` },
+                locations: [{
+                    physicalLocation: {
+                        artifactLocation,
+                        region: { startLine: x.line, startColumn: x.column, endColumn: x.column + (x.codePoint > 0xFFFF ? 2 : 1) },
+                    },
+                }],
+            });
+        }
+    }
+
+    return {
+        $schema: SARIF_SCHEMA,
+        version: '2.1.0',
+        runs: [{
+            tool: { driver: { name: 'text-sanctifier', version: VERSION, semanticVersion: VERSION, informationUri: INFORMATION_URI, rules } },
+            originalUriBaseIds: { '%SRCROOT%': { uri: pathToFileURL(cwd.endsWith(sep) ? cwd : cwd + sep).href } },
+            columnKind: 'utf16CodeUnits',
+            invocations: [{ executionSuccessful: succeeded }],
+            results,
+            properties: {
+                total,
+                filesScanned,
+                filesWithFindings: files.length,
+                truncated: files.filter((f) => f.truncated).map((f) => f.path),
+                skipped,
+            },
+        }],
+    };
+}
+
 
 // --- Main ---
 
@@ -384,7 +590,9 @@ async function main(argv) {
 
     const cwd = process.cwd();
     let paths = opts.paths;
-    if (paths.length === 0) {
+    if (paths.length === 0 && opts.aiConfig) {
+        paths = ['.'];
+    } else if (paths.length === 0) {
         if (process.stdin.isTTY) {
             err('text-sanctifier: no paths given and nothing piped on stdin.\nRun "text-sanctifier --help" for usage.');
             return 2;
@@ -393,7 +601,7 @@ async function main(argv) {
     }
 
     const warn = (msg) => err(`text-sanctifier: ${msg}`);
-    const { inputs, errors: pathErrors } = collectInputs(paths, cwd, warn);
+    const { inputs, errors: pathErrors } = collectInputs(paths, cwd, warn, opts.aiConfig);
     let errors = pathErrors;
     /** @type {Array<{ path: string, reason: string }>} */
     const skipped = [];
@@ -428,23 +636,49 @@ async function main(argv) {
         let scanned = 0;
         let total = 0;
 
+        let hiddenRunCount = 0;
+        /** @type {string[]} */
+        const lines = [];
+        const findingLine = (label, x) => `${label}:${x.line}:${x.column}  ${hex(x.codePoint)}  ${x.category}`;
+
         await forEachText((label, text) => {
             scanned++;
-            const report = inspect(text, opts.maxFindings, opts.stripBom);
+            if (!opts.aiConfig) {
+                const report = inspect(text, opts.maxFindings, opts.stripBom);
+                total += report.total;
+                if (report.total === 0) return;
+                files.push({ path: label, ...report });
+                for (const x of report.findings) lines.push(findingLine(label, x));
+                if (report.truncated) lines.push(`${label}: ${report.total - report.findings.length} more not listed (--max-findings)`);
+                return;
+            }
+
+            const { report, hiddenRuns, loose } = inspectAiConfig(text, opts.maxFindings, opts.stripBom);
             total += report.total;
-            if (report.total > 0) files.push({ path: label, ...report });
+            hiddenRunCount += report.hiddenRunCount;
+            if (report.total === 0) return;
+            files.push({ path: label, ...report });
+            // Text output lists each run once, in place of its characters
+            const entries = [
+                ...hiddenRuns.map((run) => ({ offset: run.offset, text: `${label}:${run.line}:${run.column}  ${hex(run.codePoint)}  ${describeRun(run)}` })),
+                ...loose.map((x) => ({ offset: x.offset, text: findingLine(label, x) })),
+            ].sort((a, b) => a.offset - b.offset);
+            for (const e of entries.slice(0, opts.maxFindings)) lines.push(e.text);
+            if (entries.length > opts.maxFindings) lines.push(`${label}: ${entries.length - opts.maxFindings} more not listed (--max-findings)`);
         });
 
-        if (opts.json) {
-            out(JSON.stringify({ files, total, filesScanned: scanned, filesWithFindings: files.length, skipped }, null, 2));
+        const kind = opts.aiConfig ? 'AI configuration file' : 'file';
+        if (opts.format === 'json') {
+            const summary = { files, total, filesScanned: scanned, filesWithFindings: files.length, skipped };
+            if (opts.aiConfig) summary.hiddenRunCount = hiddenRunCount;
+            out(JSON.stringify(summary, null, 2));
+        } else if (opts.format === 'sarif') {
+            out(JSON.stringify(sarifLog(files, { cwd, total, filesScanned: scanned, skipped, succeeded: errors === 0 }), null, 2));
         } else {
-            for (const f of files) {
-                for (const x of f.findings) out(`${f.path}:${x.line}:${x.column}  ${hex(x.codePoint)}  ${x.category}`);
-                if (f.truncated) out(`${f.path}: ${f.total - f.findings.length} more not listed (--max-findings)`);
-            }
-            out(total === 0
-                ? `No invisible or control characters in ${scanned} file${scanned === 1 ? '' : 's'}.`
-                : `${total} finding${total === 1 ? '' : 's'} in ${files.length} of ${scanned} file${scanned === 1 ? '' : 's'}.`);
+            for (const line of lines) out(line);
+            if (opts.aiConfig && scanned === 0 && skipped.length === 0) out('No AI configuration files found.');
+            else if (total === 0) out(`No ${opts.aiConfig ? 'hidden text, ' : ''}invisible or control characters in ${plural(scanned, kind)}.`);
+            else out(`${plural(total, 'finding')}${opts.aiConfig ? ` (${plural(hiddenRunCount, 'hidden run')})` : ''} in ${files.length} of ${plural(scanned, kind)}.`);
         }
         return errors > 0 ? 2 : total > 0 ? 1 : 0;
     }
@@ -493,7 +727,7 @@ async function main(argv) {
         files.push({ path: label, removed, removedTotal, newlinesNormalized });
     });
 
-    if (opts.json) {
+    if (opts.format === 'json') {
         report(JSON.stringify({ files, dryRun: opts.dryRun, filesScanned: scanned, filesChanged: files.length, skipped }, null, 2));
     } else {
         const verb = opts.dryRun ? 'would fix' : 'fixed';

package/dist/text-sanctifier.min.js

@@ -1,14 +1,15 @@
-function n(a={}){const c=!!a.purgeInvisibleChars,d=!!a.purgeEmojis,e=!!a.nukeControls,g=!!a.keyboardOnlyFilter,f=!!a.normalizeNewlines,l=!!a.trimSpacesAroundNewlines,b=!!a.collapseNewLines,h=!!a.preserveParagraphs,k=!!a.collapseSpaces,r=!!a.finalTrim,m=Number.isFinite(a.maxLength)&&0<=a.maxLength?a.maxLength:Infinity,p="truncate"===a.g||"noop"===a.g||"throw"===a.g?a.g:"throw";return A=>q(A,c,d,e,g,f,l,b,h,k,r,m,p)}n.strict=a=>q(a,!0,!0,!0,!1,!0,!0,!0,!1,!0,!0);
-n.loose=a=>q(a,!1,!1,!1,!1,!0,!0,!0,!0,!0,!0);n.keyboardOnlyEmoji=a=>q(a,!1,!1,!1,!0,!0,!0,!1,!1,!1,!0);n.keyboardOnly=a=>q(a,!0,!0,!0,!0,!0,!0,!0,!1,!0,!0);
-function q(a,c=!1,d=!1,e=!1,g=!1,f=!1,l=!1,b=!1,h=!1,k=!1,r=!1,m=Infinity,p="throw"){if("string"!==typeof a)throw new TypeError("sanctifyText expects a string input.");"truncate"!==p&&"noop"!==p&&"throw"!==p&&(p="throw");if(a.length>m)switch(p){case "truncate":a=a.slice(0,m);break;case "noop":return a;default:throw new RangeError(`sanctifyText input length ${a.length} exceeds maxLength ${m}.`);}c&&(a=a.replace(t,""));d&&(a=a.replace(u,""));e&&(a=a.replace(v,""));g&&(a=w(a,d));f&&(a=a.replace(x,"\n"));
-l&&(a=a.replace(y,"$1"));b&&(c=a,a=h?c.replace(z,"\n\n"):c.replace(B,"\n"));k&&(a=a.replace(C," "));return r?a.trim():a}var t=/[\u00a0\u00ad\u034f\u061c\u115f\u1160\u180e\u2000-\u200f\u202a-\u202f\u205f\u2060-\u2064\u2066-\u2069\u3000\u3164\ufe00-\ufe0f\ufeff\uffa0\u{E0000}-\u{E007F}\u{E0100}-\u{E01EF}]+/gu,D=/[^\x20-\x7E\r\n]+/gu;
-function w(a,c=!1){a=E(a);return c?a.replace(D,""):a.replace(D,d=>{u.lastIndex=0;if(u.test(d)){u.lastIndex=0;var e="";for(const g of d.matchAll(u))e+=g[0];d=e}else d="";return d})}var F=/[\u2018\u2019\u201A\u201B\u2032\u2035]/g,G=/[\u201C\u201D\u201E\u201F\u2033\u2036\u00AB\u00BB]/g,H=/[\u2012\u2013\u2014\u2015\u2212]/g,I=/\u2026/g,J=/[\u2022\u00B7]/g,K=/[\uFF01-\uFF5E]/g;
-function E(a){return a.replace(F,"'").replace(G,'"').replace(H,"-").replace(I,"...").replace(J,"*").replace(K,c=>String.fromCharCode(c.charCodeAt(0)-65248))}var u;try{u=RegExp("(?:\\p{Extended_Pictographic}(?:\\uFE0F|\\uFE0E)?(?:\\u200D(?:\\p{Extended_Pictographic}|\\w)+)*)","gu")}catch{u=/[\u{2700}-\u{27BF}\u{1F300}-\u{1FAFF}\u{1F3FB}-\u{1F3FF}\u200D\uFE0F\u{1F1E6}-\u{1F1FF}]/gu}var x=/\r\n|\r|\n/g,y=/[ \t]*(\n+)[ \t]*/g,B=/\n{2,}/g,z=/\n{3,}/g,C=/ {2,}/g,v=/[\u0000-\u0008\u000B\u000C\u000E-\u001F\u007F\u0080-\u009F\u061C\u200E\u200F\u202A-\u202E\u2066-\u2069]+/g;
-function L(a,c){a.lastIndex=0;return a.test(c)}
-function M(a){if("string"!==typeof a)throw new TypeError("inspectText expects a string input.");const c=[],d={hasControlChars:!1,hasInvisibleChars:!1,hasMixedNewlines:!1,newlineStyle:null,hasEmojis:!1,hasNonKeyboardChars:!1,summary:c};L(v,a)&&(d.hasControlChars=!0,c.push("Control characters detected."));L(t,a)&&(d.hasInvisibleChars=!0,c.push("Invisible Unicode characters detected."));L(u,a)&&(d.hasEmojis=!0,c.push("Emojis detected."));const {mixed:e,types:g}=N(a);d.hasMixedNewlines=e;d.newlineStyle=
-e?"Mixed":g[0]||null;d.newlineStyle&&c.push(e?"Mixed newline styles detected.":`Consistent newline style: ${d.newlineStyle}`);a:{a=E(a);if(L(D,a))for(f of a)if(!("\n"===f||"\r"===f||" "<=f&&"~">=f||L(u,f))){var f=!0;break a}f=!1}f&&(d.hasNonKeyboardChars=!0,c.push("Non-keyboard characters detected."));return d}
-function O(a,c={}){if("string"!==typeof a)throw new TypeError("inspectTextDetailed expects a string input.");c=c&&c.maxFindings;c=Infinity===c||Number.isInteger(c)&&0<=c?c:100;const d={bidi_control:0,zero_width:0,nonbreaking_space:0,bom:0,nul:0,c0_control:0,c1_control:0,invisible_format:0,hangul_filler:0,variation_selector:0,tag:0},e=[];let g=0,f=1,l=0;for(let k=0,r=a.length;k<r;k++){var b=a.charCodeAt(k);if(32<=b&&127>b)continue;if(10===b){f++;l=k+1;continue}if(13===b){10!==a.charCodeAt(k+1)&&(f++,
-l=k+1);continue}const m=k;var h=b;if(56128===b){h=a.charCodeAt(k+1);if(!(56320<=h&&57343>=h))continue;h=65536+(b-55296<<10)+(h-56320);b=917504<=h&&917631>=h?"tag":917760<=h&&917999>=h?"variation_selector":null;k++}else b=128>b?0===b?"nul":32>b?9===b||10===b||13===b?null:"c0_control":127===b?"c0_control":null:159>=b?"c1_control":160===b?"nonbreaking_space":173===b||847===b||6158===b?"invisible_format":1564===b?"bidi_control":4447===b||4448===b?"hangul_filler":8192>b?null:8202>=b?"nonbreaking_space":
-8205>=b?"zero_width":8207>=b||8234<=b&&8238>=b?"bidi_control":8239===b||8287===b?"nonbreaking_space":8288===b?"zero_width":8289<=b&&8292>=b?"invisible_format":8294<=b&&8297>=b?"bidi_control":12288===b?"nonbreaking_space":12644===b||65440===b?"hangul_filler":65024<=b&&65039>=b?"variation_selector":65279===b?"bom":null;null!==b&&(d[b]++,g++,e.length<c&&(h={offset:m,line:f,column:m-l+1,codePoint:h,category:b},"bom"===b&&(h.atStart=0===m),e.push(h)))}return{findings:e,counts:d,total:g,truncated:g>e.length}}
-function N(a){if("string"!==typeof a)throw new TypeError("getNewlineStats expects a string input.");var c=0;let d=0,e=0;for(let g=0,f=a.length;g<f;g++){const l=a.charCodeAt(g);13===l?10===a.charCodeAt(g+1)?(c++,g++):d++:10===l&&e++}a={crlf:c,cr:d,lf:e};c=[];0<a.crlf&&c.push("CRLF");0<a.cr&&c.push("CR");0<a.lf&&c.push("LF");return{...a,types:c,mixed:1<c.length}}
-function P(a){return{purgeInvisibleChars:a.hasInvisibleChars,purgeEmojis:a.hasEmojis,nukeControls:a.hasControlChars,keyboardOnlyFilter:a.hasNonKeyboardChars,normalizeNewlines:a.hasMixedNewlines||"CRLF"===a.newlineStyle||"CR"===a.newlineStyle}}export { n as summonSanctifier, M as inspectText, P as getRecommendedSanctifierOptions, O as inspectTextDetailed, N as getNewlineStats };
+var n=/[\u0000-\u0008\u000B\u000C\u000E-\u001F\u007F-\u009F\u061C\u200E\u200F\u202A-\u202E\u2066-\u2069]+/g,q=/[\u00A0\u00AD\u034F\u061C\u115F\u1160\u17B4\u17B5\u180B-\u180F\u2000-\u200F\u202A-\u202F\u205F-\u206F\u3000\u3164\uFE00-\uFE0F\uFEFF\uFFA0\uFFF0-\uFFF8\u{1BCA0}-\u{1BCA3}\u{1D173}-\u{1D17A}\u{E0000}-\u{E0FFF}]+/gu;
+function t(a={}){const c=!!a.purgeInvisibleChars,d=!!a.purgeEmojis,f=!!a.nukeControls,h=!!a.keyboardOnlyFilter,g=!!a.normalizeNewlines,l=!!a.trimSpacesAroundNewlines,b=!!a.collapseNewLines,e=!!a.preserveParagraphs,k=!!a.collapseSpaces,r=!!a.finalTrim,m=Number.isFinite(a.maxLength)&&0<=a.maxLength?a.maxLength:Infinity,p="truncate"===a.g||"noop"===a.g||"throw"===a.g?a.g:"throw";return A=>u(A,c,d,f,h,g,l,b,e,k,r,m,p)}t.strict=a=>u(a,!0,!0,!0,!1,!0,!0,!0,!1,!0,!0);
+t.loose=a=>u(a,!1,!1,!1,!1,!0,!0,!0,!0,!0,!0);t.keyboardOnlyEmoji=a=>u(a,!1,!1,!1,!0,!0,!0,!1,!1,!1,!0);t.keyboardOnly=a=>u(a,!0,!0,!0,!0,!0,!0,!0,!1,!0,!0);
+function u(a,c=!1,d=!1,f=!1,h=!1,g=!1,l=!1,b=!1,e=!1,k=!1,r=!1,m=Infinity,p="throw"){if("string"!==typeof a)throw new TypeError("sanctifyText expects a string input.");"truncate"!==p&&"noop"!==p&&"throw"!==p&&(p="throw");if(a.length>m)switch(p){case "truncate":a=a.slice(0,m);break;case "noop":return a;default:throw new RangeError(`sanctifyText input length ${a.length} exceeds maxLength ${m}.`);}c&&(a=a.replace(q,""));d&&(a=a.replace(v,""));f&&(a=a.replace(n,""));h&&(a=w(a,d));g&&(a=a.replace(x,"\n"));
+l&&(a=a.replace(y,"$1"));b&&(c=a,a=e?c.replace(z,"\n\n"):c.replace(B,"\n"));k&&(a=a.replace(C," "));return r?a.trim():a}var D=/[^\x20-\x7E\r\n]+/gu;function w(a,c=!1){a=E(a);return c?a.replace(D,""):a.replace(D,d=>{v.lastIndex=0;if(v.test(d)){v.lastIndex=0;var f="";for(const h of d.matchAll(v))f+=h[0];d=f}else d="";return d})}
+var F=/[\u2018\u2019\u201A\u201B\u2032\u2035]/g,G=/[\u201C\u201D\u201E\u201F\u2033\u2036\u00AB\u00BB]/g,H=/[\u2012\u2013\u2014\u2015\u2212]/g,I=/\u2026/g,J=/[\u2022\u00B7]/g,K=/[\uFF01-\uFF5E]/g;function E(a){return a.replace(F,"'").replace(G,'"').replace(H,"-").replace(I,"...").replace(J,"*").replace(K,c=>String.fromCharCode(c.charCodeAt(0)-65248))}var v;try{v=RegExp("(?:\\p{Extended_Pictographic}(?:\\uFE0F|\\uFE0E)?(?:\\u200D(?:\\p{Extended_Pictographic}|\\w)+)*)","gu")}catch{v=/[\u{2700}-\u{27BF}\u{1F300}-\u{1FAFF}\u{1F3FB}-\u{1F3FF}\u200D\uFE0F\u{1F1E6}-\u{1F1FF}]/gu}
+var x=/\r\n|\r|\n/g,y=/[ \t]*(\n+)[ \t]*/g,B=/\n{2,}/g,z=/\n{3,}/g,C=/ {2,}/g;function L(a,c){a.lastIndex=0;return a.test(c)}
+function M(a){if("string"!==typeof a)throw new TypeError("inspectText expects a string input.");const c=[],d={hasControlChars:!1,hasInvisibleChars:!1,hasMixedNewlines:!1,newlineStyle:null,hasEmojis:!1,hasNonKeyboardChars:!1,summary:c};L(n,a)&&(d.hasControlChars=!0,c.push("Control characters detected."));L(q,a)&&(d.hasInvisibleChars=!0,c.push("Invisible Unicode characters detected."));L(v,a)&&(d.hasEmojis=!0,c.push("Emojis detected."));const {mixed:f,types:h}=N(a);d.hasMixedNewlines=f;d.newlineStyle=
+f?"Mixed":h[0]||null;d.newlineStyle&&c.push(f?"Mixed newline styles detected.":`Consistent newline style: ${d.newlineStyle}`);a:{a=E(a);if(L(D,a))for(g of a)if(!("\n"===g||"\r"===g||" "<=g&&"~">=g||L(v,g))){var g=!0;break a}g=!1}g&&(d.hasNonKeyboardChars=!0,c.push("Non-keyboard characters detected."));return d}
+function O(a,c={}){if("string"!==typeof a)throw new TypeError("inspectTextDetailed expects a string input.");c=c&&c.maxFindings;c=Infinity===c||Number.isInteger(c)&&0<=c?c:100;const d={bidi_control:0,zero_width:0,nonbreaking_space:0,bom:0,nul:0,c0_control:0,c1_control:0,invisible_format:0,hangul_filler:0,variation_selector:0,tag:0},f=[];let h=0,g=1,l=0;for(let k=0,r=a.length;k<r;k++){var b=a.charCodeAt(k);if(32<=b&&127>b)continue;if(10===b){g++;l=k+1;continue}if(13===b){10!==a.charCodeAt(k+1)&&(g++,
+l=k+1);continue}const m=k;var e=b;if(55296<=b&&56319>=b){e=a.charCodeAt(k+1);if(!(56320<=e&&57343>=e))continue;e=65536+(b-55296<<10)+(e-56320);b=917504>e?113828>e?113824>e?null:"invisible_format":119155>e?null:119163>e?"invisible_format":null:917760>e?917632>e?"tag":"invisible_format":918E3>e?"variation_selector":921600>e?"invisible_format":null;k++}else b=8192>b?174>b?32>b?11>b?1>b?"nul":9>b?"c0_control":null:13>b?"c0_control":14>b?null:"c0_control":160>b?127>b?null:128>b?"c0_control":"c1_control":
+161>b?"nonbreaking_space":173>b?null:"invisible_format":4449>b?1564>b?847>b?null:848>b?"invisible_format":null:1565>b?"bidi_control":4447>b?null:"hangul_filler":6155>b?6068>b?null:6070>b?"invisible_format":null:6159>b?6158>b?"variation_selector":"invisible_format":6160>b?"variation_selector":null:8304>b?8240>b?8208>b?8203>b?"nonbreaking_space":8206>b?"zero_width":"bidi_control":8234>b?null:8239>b?"bidi_control":"nonbreaking_space":8289>b?8287>b?null:8288>b?"nonbreaking_space":"zero_width":8294>b?
+"invisible_format":8298>b?"bidi_control":"invisible_format":65040>b?12644>b?12288>b?null:12289>b?"nonbreaking_space":null:12645>b?"hangul_filler":65024>b?null:"variation_selector":65440>b?65279>b?null:65280>b?"bom":null:65520>b?65441>b?"hangul_filler":null:65529>b?"invisible_format":null;null!==b&&(d[b]++,h++,f.length<c&&(e={offset:m,line:g,column:m-l+1,codePoint:e,category:b},"bom"===b&&(e.atStart=0===m),f.push(e)))}return{findings:f,counts:d,total:h,truncated:h>f.length}}
+function N(a){if("string"!==typeof a)throw new TypeError("getNewlineStats expects a string input.");var c=0;let d=0,f=0;for(let h=0,g=a.length;h<g;h++){const l=a.charCodeAt(h);13===l?10===a.charCodeAt(h+1)?(c++,h++):d++:10===l&&f++}a={crlf:c,cr:d,lf:f};c=[];0<a.crlf&&c.push("CRLF");0<a.cr&&c.push("CR");0<a.lf&&c.push("LF");return{...a,types:c,mixed:1<c.length}}
+function P(a){return{purgeInvisibleChars:a.hasInvisibleChars,purgeEmojis:a.hasEmojis,nukeControls:a.hasControlChars,keyboardOnlyFilter:a.hasNonKeyboardChars,normalizeNewlines:a.hasMixedNewlines||"CRLF"===a.newlineStyle||"CR"===a.newlineStyle}}export { t as summonSanctifier, M as inspectText, P as getRecommendedSanctifierOptions, O as inspectTextDetailed, N as getNewlineStats };