@postalsys/email-content 0.0.1

package/dist/util/lang.js

@@ -0,0 +1,135 @@
+"use strict";
+/**
+ * Language estimation.
+ *
+ * Deliberately small and dependency free. This is an estimate used for stratified sampling and for
+ * reporting metrics per language, never for a classification decision. A wrong guess costs a
+ * slightly skewed sample, nothing more, which is why a stopword scorer is preferred over pulling in
+ * a model.
+ *
+ * Two stages: script detection first, because a Cyrillic or CJK body is unambiguous and cheap to
+ * spot, then stopword frequency for Latin script languages.
+ */
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.estimateLanguage = estimateLanguage;
+/**
+ * Stopword lists. Kept short on purpose: the most frequent function words carry nearly all of the
+ * discriminative power, and long lists mostly add overlap between related languages.
+ */
+const STOPWORDS = {
+    en: [
+        'the',
+        'and',
+        'you',
+        'that',
+        'for',
+        'with',
+        'this',
+        'have',
+        'not',
+        'are',
+        'was',
+        'but',
+        'from',
+        'they',
+        'will',
+        'would',
+        'there',
+        'their',
+        'what',
+        'about'
+    ],
+    et: ['ja', 'on', 'ei', 'et', 'see', 'ka', 'kui', 'aga', 'mis', 'oma', 'ning', 'siis', 'ole', 'seda', 'kas', 'nii', 'veel', 'ta', 'ma', 'või'],
+    de: ['der', 'die', 'und', 'das', 'ist', 'nicht', 'ich', 'sie', 'mit', 'den', 'von', 'für', 'auf', 'wir', 'auch', 'dem', 'ein', 'eine', 'aber', 'noch'],
+    fr: ['les', 'des', 'est', 'une', 'que', 'pour', 'dans', 'pas', 'vous', 'sur', 'avec', 'nous', 'plus', 'par', 'mais', 'ils', 'ce', 'au', 'du', 'je'],
+    es: ['que', 'los', 'las', 'del', 'por', 'con', 'para', 'una', 'como', 'más', 'pero', 'sus', 'este', 'esta', 'son', 'muy', 'ser', 'sin', 'sobre', 'todo'],
+    it: ['che', 'per', 'con', 'del', 'non', 'una', 'sono', 'della', 'come', 'più', 'anche', 'nel', 'alla', 'ma', 'si', 'da', 'gli', 'questo', 'essere', 'ho'],
+    pt: [
+        'que',
+        'não',
+        'com',
+        'para',
+        'uma',
+        'dos',
+        'das',
+        'por',
+        'mais',
+        'como',
+        'mas',
+        'foi',
+        'ele',
+        'ser',
+        'você',
+        'está',
+        'pelo',
+        'isso',
+        'quando',
+        'muito'
+    ],
+    nl: ['het', 'een', 'van', 'de', 'is', 'dat', 'niet', 'en', 'op', 'te', 'zijn', 'voor', 'met', 'maar', 'ook', 'aan', 'er', 'dan', 'wij', 'heeft'],
+    sv: ['och', 'att', 'det', 'som', 'en', 'är', 'på', 'för', 'med', 'den', 'har', 'inte', 'om', 'jag', 'men', 'de', 'till', 'kan', 'vi', 'ett'],
+    fi: ['on', 'ja', 'ei', 'se', 'että', 'ovat', 'ole', 'kun', 'niin', 'mutta', 'voi', 'tai', 'jos', 'sen', 'me', 'te', 'hän', 'kuin', 'vain', 'myös'],
+    pl: ['nie', 'się', 'jest', 'że', 'do', 'na', 'to', 'jak', 'ale', 'dla', 'przez', 'tego', 'lub', 'oraz', 'tym', 'przy', 'być', 'ma', 'już', 'tylko']
+};
+const SCRIPT_RANGES = [
+    { lang: 'ru', regex: /[Ѐ-ӿ]/g },
+    { lang: 'el', regex: /[Ͱ-Ͽ]/g },
+    { lang: 'he', regex: /[֐-׿]/g },
+    { lang: 'ar', regex: /[؀-ۿ]/g },
+    { lang: 'ko', regex: /[가-힯ᄀ-ᇿ]/g },
+    { lang: 'ja', regex: /[぀-ヿ]/g },
+    { lang: 'zh', regex: /[一-鿿]/g }
+];
+/** Minimum tokens needed before a Latin script guess is trusted. */
+const MIN_TOKENS = 12;
+/** Minimum matched share before a guess is returned rather than "unknown". */
+const MIN_SCORE = 0.06;
+/**
+ * Estimate the language of a body of text.
+ *
+ * @param text plain text. Only the first 20k characters are inspected, which is ample and keeps the
+ *   cost bounded on newsletters and digests.
+ */
+function estimateLanguage(text) {
+    const sample = text.length > 20000 ? text.slice(0, 20000) : text;
+    if (!sample.trim()) {
+        return { lang: 'unknown', score: 0 };
+    }
+    // Script detection first. A body with a meaningful share of non Latin letters is decided here.
+    const letters = sample.replace(/[^\p{L}]/gu, '').length;
+    if (letters > 20) {
+        for (const { lang, regex } of SCRIPT_RANGES) {
+            regex.lastIndex = 0;
+            const matches = sample.match(regex);
+            if (matches && matches.length / letters > 0.2) {
+                return { lang, score: matches.length / letters };
+            }
+        }
+    }
+    const tokens = sample
+        .toLowerCase()
+        .split(/[^\p{L}']+/u)
+        .filter(token => token.length > 1);
+    if (tokens.length < MIN_TOKENS) {
+        return { lang: 'unknown', score: 0 };
+    }
+    const counts = new Map();
+    for (const token of tokens) {
+        counts.set(token, (counts.get(token) || 0) + 1);
+    }
+    let bestLang = 'unknown';
+    let bestScore = 0;
+    for (const [lang, words] of Object.entries(STOPWORDS)) {
+        let hits = 0;
+        for (const word of words) {
+            hits += counts.get(word) || 0;
+        }
+        const score = hits / tokens.length;
+        if (score > bestScore) {
+            bestScore = score;
+            bestLang = lang;
+        }
+    }
+    return bestScore >= MIN_SCORE ? { lang: bestLang, score: bestScore } : { lang: 'unknown', score: bestScore };
+}
+//# sourceMappingURL=lang.js.map

package/dist/util/lang.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"lang.js","sourceRoot":"","sources":["../../src/util/lang.ts"],"names":[],"mappings":";AAAA;;;;;;;;;;GAUG;;AA6FH,4CA8CC;AAhID;;;GAGG;AACH,MAAM,SAAS,GAAsC;IACjD,EAAE,EAAE;QACA,KAAK;QACL,KAAK;QACL,KAAK;QACL,MAAM;QACN,KAAK;QACL,MAAM;QACN,MAAM;QACN,MAAM;QACN,KAAK;QACL,KAAK;QACL,KAAK;QACL,KAAK;QACL,MAAM;QACN,MAAM;QACN,MAAM;QACN,OAAO;QACP,OAAO;QACP,OAAO;QACP,MAAM;QACN,OAAO;KACV;IACD,EAAE,EAAE,CAAC,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,KAAK,EAAE,IAAI,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,IAAI,EAAE,IAAI,EAAE,KAAK,CAAC;IAC7I,EAAE,EAAE,CAAC,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,OAAO,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,MAAM,EAAE,MAAM,CAAC;IACtJ,EAAE,EAAE,CAAC,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,MAAM,EAAE,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,KAAK,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,CAAC;IACnJ,EAAE,EAAE,CAAC,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,MAAM,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,OAAO,EAAE,MAAM,CAAC;IACxJ,EAAE,EAAE,CAAC,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,OAAO,EAAE,MAAM,EAAE,KAAK,EAAE,OAAO,EAAE,KAAK,EAAE,MAAM,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,KAAK,EAAE,QAAQ,EAAE,QAAQ,EAAE,IAAI,CAAC;IACzJ,EAAE,EAAE;QACA,KAAK;QACL,KAAK;QACL,KAAK;QACL,MAAM;QACN,KAAK;QACL,KAAK;QACL,KAAK;QACL,KAAK;QACL,MAAM;QACN,MAAM;QACN,KAAK;QACL,KAAK;QACL,KAAK;QACL,KAAK;QACL,MAAM;QACN,MAAM;QACN,MAAM;QACN,MAAM;QACN,QAAQ;QACR,OAAO;KACV;IACD,EAAE,EAAE,CAAC,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,IAAI,EAAE,IAAI,EAAE,KAAK,EAAE,MAAM,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,MAAM,EAAE,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,KAAK,EAAE,KAAK,EAAE,IAAI,EAAE,KAAK,EAAE,KAAK,EAAE,OAAO,CAAC;IAChJ,EAAE,EAAE,CAAC,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,IAAI,EAAE,KAAK,EAAE,KAAK,EAAE,IAAI,EAAE,MAAM,EAAE,KAAK,EAAE,IAAI,EAAE,KAAK,CAAC;IAC5I,EAAE,EAAE,CAAC,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,MAAM,EAAE,MAAM,EAAE,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,OAAO,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,IAAI,EAAE,IAAI,EAAE,KAAK,EAAE,MAAM,EAAE,MAAM,EAAE,MAAM,CAAC;IAClJ,EAAE,EAAE,CAAC,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,OAAO,EAAE,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,KAAK,EAAE,IAAI,EAAE,KAAK,EAAE,OAAO,CAAC;CACtJ,CAAC;AAEF,MAAM,aAAa,GAAqD;IACpE,EAAE,IAAI,EAAE,IAAI,EAAE,KAAK,EAAE,QAAQ,EAAE;IAC/B,EAAE,IAAI,EAAE,IAAI,EAAE,KAAK,EAAE,QAAQ,EAAE;IAC/B,EAAE,IAAI,EAAE,IAAI,EAAE,KAAK,EAAE,QAAQ,EAAE;IAC/B,EAAE,IAAI,EAAE,IAAI,EAAE,KAAK,EAAE,QAAQ,EAAE;IAC/B,EAAE,IAAI,EAAE,IAAI,EAAE,KAAK,EAAE,WAAW,EAAE;IAClC,EAAE,IAAI,EAAE,IAAI,EAAE,KAAK,EAAE,QAAQ,EAAE;IAC/B,EAAE,IAAI,EAAE,IAAI,EAAE,KAAK,EAAE,QAAQ,EAAE;CAClC,CAAC;AAEF,oEAAoE;AACpE,MAAM,UAAU,GAAG,EAAE,CAAC;AAEtB,8EAA8E;AAC9E,MAAM,SAAS,GAAG,IAAI,CAAC;AAEvB;;;;;GAKG;AACH,SAAgB,gBAAgB,CAAC,IAAY;IACzC,MAAM,MAAM,GAAG,IAAI,CAAC,MAAM,GAAG,KAAK,CAAC,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,EAAE,KAAK,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC;IACjE,IAAI,CAAC,MAAM,CAAC,IAAI,EAAE,EAAE,CAAC;QACjB,OAAO,EAAE,IAAI,EAAE,SAAS,EAAE,KAAK,EAAE,CAAC,EAAE,CAAC;IACzC,CAAC;IAED,+FAA+F;IAC/F,MAAM,OAAO,GAAG,MAAM,CAAC,OAAO,CAAC,YAAY,EAAE,EAAE,CAAC,CAAC,MAAM,CAAC;IACxD,IAAI,OAAO,GAAG,EAAE,EAAE,CAAC;QACf,KAAK,MAAM,EAAE,IAAI,EAAE,KAAK,EAAE,IAAI,aAAa,EAAE,CAAC;YAC1C,KAAK,CAAC,SAAS,GAAG,CAAC,CAAC;YACpB,MAAM,OAAO,GAAG,MAAM,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC;YACpC,IAAI,OAAO,IAAI,OAAO,CAAC,MAAM,GAAG,OAAO,GAAG,GAAG,EAAE,CAAC;gBAC5C,OAAO,EAAE,IAAI,EAAE,KAAK,EAAE,OAAO,CAAC,MAAM,GAAG,OAAO,EAAE,CAAC;YACrD,CAAC;QACL,CAAC;IACL,CAAC;IAED,MAAM,MAAM,GAAG,MAAM;SAChB,WAAW,EAAE;SACb,KAAK,CAAC,aAAa,CAAC;SACpB,MAAM,CAAC,KAAK,CAAC,EAAE,CAAC,KAAK,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;IACvC,IAAI,MAAM,CAAC,MAAM,GAAG,UAAU,EAAE,CAAC;QAC7B,OAAO,EAAE,IAAI,EAAE,SAAS,EAAE,KAAK,EAAE,CAAC,EAAE,CAAC;IACzC,CAAC;IAED,MAAM,MAAM,GAAG,IAAI,GAAG,EAAkB,CAAC;IACzC,KAAK,MAAM,KAAK,IAAI,MAAM,EAAE,CAAC;QACzB,MAAM,CAAC,GAAG,CAAC,KAAK,EAAE,CAAC,MAAM,CAAC,GAAG,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC;IACpD,CAAC;IAED,IAAI,QAAQ,GAAiB,SAAS,CAAC;IACvC,IAAI,SAAS,GAAG,CAAC,CAAC;IAClB,KAAK,MAAM,CAAC,IAAI,EAAE,KAAK,CAAC,IAAI,MAAM,CAAC,OAAO,CAAC,SAAS,CAAC,EAAE,CAAC;QACpD,IAAI,IAAI,GAAG,CAAC,CAAC;QACb,KAAK,MAAM,IAAI,IAAI,KAAK,EAAE,CAAC;YACvB,IAAI,IAAI,MAAM,CAAC,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QAClC,CAAC;QACD,MAAM,KAAK,GAAG,IAAI,GAAG,MAAM,CAAC,MAAM,CAAC;QACnC,IAAI,KAAK,GAAG,SAAS,EAAE,CAAC;YACpB,SAAS,GAAG,KAAK,CAAC;YAClB,QAAQ,GAAG,IAAoB,CAAC;QACpC,CAAC;IACL,CAAC;IAED,OAAO,SAAS,IAAI,SAAS,CAAC,CAAC,CAAC,EAAE,IAAI,EAAE,QAAQ,EAAE,KAAK,EAAE,SAAS,EAAE,CAAC,CAAC,CAAC,EAAE,IAAI,EAAE,SAAS,EAAE,KAAK,EAAE,SAAS,EAAE,CAAC;AACjH,CAAC"}

package/dist/util/worker-pool.d.ts

@@ -0,0 +1,51 @@
+/**
+ * Fixed size worker pool with per task timeouts.
+ *
+ * Adapted from the pool in @postalsys/email-text-tools (lib/worker-pool.js), simplified for batch
+ * work: the corpus scanner knows how many workers it wants and drives its own backpressure, so the
+ * elastic sizing and idle reaping of the original are not needed. What is kept is the important
+ * part, a task timeout that terminates and replaces the worker, so that one pathological message
+ * cannot wedge a scan of 320k files.
+ */
+import { EventEmitter } from 'node:events';
+export interface WorkerPoolOptions {
+    workerPath: string;
+    size: number;
+    /** Per task timeout in milliseconds. */
+    taskTimeoutMs: number;
+    /** Optional data passed to every worker at startup. */
+    workerData?: unknown;
+    /**
+     * Old generation heap cap per worker, in MiB.
+     *
+     * Worker threads share the parent process, so their heaps all count towards its RSS. Without a
+     * cap, sixteen workers each growing to a comfortable default is more than a gigabyte of
+     * resident memory for a job whose working set is one message at a time. Exceeding the cap
+     * terminates that worker, which the pool treats like any other worker failure.
+     */
+    maxOldGenerationSizeMb?: number;
+}
+export declare class WorkerTimeoutError extends Error {
+    constructor(ms: number);
+}
+/**
+ * Emits 'worker-error' with the underlying error whenever a worker dies unexpectedly. Callers may
+ * log it; the pool itself recovers by spawning a replacement.
+ */
+export declare class WorkerPool<TTask, TResult> extends EventEmitter {
+    private readonly options;
+    private readonly slots;
+    private readonly queue;
+    private nextId;
+    private closing;
+    constructor(options: WorkerPoolOptions);
+    private spawn;
+    private replace;
+    private dispatch;
+    private drainQueue;
+    /** Number of tasks accepted but not yet completed. */
+    get inFlight(): number;
+    submit(task: TTask): Promise<TResult>;
+    close(): Promise<void>;
+}
+//# sourceMappingURL=worker-pool.d.ts.map

package/dist/util/worker-pool.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"worker-pool.d.ts","sourceRoot":"","sources":["../../src/util/worker-pool.ts"],"names":[],"mappings":"AAAA;;;;;;;;GAQG;AAGH,OAAO,EAAE,YAAY,EAAE,MAAM,aAAa,CAAC;AAE3C,MAAM,WAAW,iBAAiB;IAC9B,UAAU,EAAE,MAAM,CAAC;IACnB,IAAI,EAAE,MAAM,CAAC;IACb,wCAAwC;IACxC,aAAa,EAAE,MAAM,CAAC;IACtB,uDAAuD;IACvD,UAAU,CAAC,EAAE,OAAO,CAAC;IACrB;;;;;;;OAOG;IACH,sBAAsB,CAAC,EAAE,MAAM,CAAC;CACnC;AA+BD,qBAAa,kBAAmB,SAAQ,KAAK;gBAC7B,EAAE,EAAE,MAAM;CAIzB;AAED;;;GAGG;AACH,qBAAa,UAAU,CAAC,KAAK,EAAE,OAAO,CAAE,SAAQ,YAAY;IACxD,OAAO,CAAC,QAAQ,CAAC,OAAO,CAAoB;IAC5C,OAAO,CAAC,QAAQ,CAAC,KAAK,CAAuB;IAC7C,OAAO,CAAC,QAAQ,CAAC,KAAK,CAAgC;IACtD,OAAO,CAAC,MAAM,CAAK;IACnB,OAAO,CAAC,OAAO,CAAS;gBAEZ,OAAO,EAAE,iBAAiB;IAQtC,OAAO,CAAC,KAAK;IA4Cb,OAAO,CAAC,OAAO;IAaf,OAAO,CAAC,QAAQ;IAkBhB,OAAO,CAAC,UAAU;IAclB,sDAAsD;IACtD,IAAI,QAAQ,IAAI,MAAM,CAErB;IAED,MAAM,CAAC,IAAI,EAAE,KAAK,GAAG,OAAO,CAAC,OAAO,CAAC;IAe/B,KAAK,IAAI,OAAO,CAAC,IAAI,CAAC;CAgB/B"}

package/dist/util/worker-pool.js

@@ -0,0 +1,155 @@
+"use strict";
+/**
+ * Fixed size worker pool with per task timeouts.
+ *
+ * Adapted from the pool in @postalsys/email-text-tools (lib/worker-pool.js), simplified for batch
+ * work: the corpus scanner knows how many workers it wants and drives its own backpressure, so the
+ * elastic sizing and idle reaping of the original are not needed. What is kept is the important
+ * part, a task timeout that terminates and replaces the worker, so that one pathological message
+ * cannot wedge a scan of 320k files.
+ */
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.WorkerPool = exports.WorkerTimeoutError = void 0;
+const node_worker_threads_1 = require("node:worker_threads");
+const node_events_1 = require("node:events");
+class WorkerTimeoutError extends Error {
+    constructor(ms) {
+        super(`worker task timed out after ${ms}ms`);
+        this.name = 'WorkerTimeoutError';
+    }
+}
+exports.WorkerTimeoutError = WorkerTimeoutError;
+/**
+ * Emits 'worker-error' with the underlying error whenever a worker dies unexpectedly. Callers may
+ * log it; the pool itself recovers by spawning a replacement.
+ */
+class WorkerPool extends node_events_1.EventEmitter {
+    options;
+    slots = [];
+    queue = [];
+    nextId = 1;
+    closing = false;
+    constructor(options) {
+        super();
+        this.options = options;
+        for (let i = 0; i < Math.max(1, options.size); i++) {
+            this.spawn();
+        }
+    }
+    spawn() {
+        const worker = new node_worker_threads_1.Worker(this.options.workerPath, {
+            workerData: this.options.workerData,
+            resourceLimits: this.options.maxOldGenerationSizeMb ? { maxOldGenerationSizeMb: this.options.maxOldGenerationSizeMb } : undefined
+        });
+        const slot = { worker, busy: false, current: null };
+        worker.on('message', (reply) => {
+            const current = slot.current;
+            if (!current || current.id !== reply.id) {
+                return;
+            }
+            clearTimeout(current.timer);
+            slot.current = null;
+            slot.busy = false;
+            if (reply.ok) {
+                current.resolve(reply.result);
+            }
+            else {
+                current.reject(new Error(reply.error || 'worker failed'));
+            }
+            this.drainQueue();
+        });
+        worker.on('error', err => {
+            this.emit('worker-error', err);
+            const current = slot.current;
+            if (current) {
+                clearTimeout(current.timer);
+                slot.current = null;
+                current.reject(err instanceof Error ? err : new Error(String(err)));
+            }
+            this.replace(slot);
+        });
+        worker.on('exit', () => {
+            if (!this.closing && this.slots.includes(slot)) {
+                this.replace(slot);
+            }
+        });
+        this.slots.push(slot);
+        return slot;
+    }
+    replace(slot) {
+        const index = this.slots.indexOf(slot);
+        if (index !== -1) {
+            this.slots.splice(index, 1);
+        }
+        slot.busy = false;
+        void slot.worker.terminate().catch(() => undefined);
+        if (!this.closing) {
+            this.spawn();
+            this.drainQueue();
+        }
+    }
+    dispatch(slot, item) {
+        slot.busy = true;
+        const timer = setTimeout(() => {
+            const current = slot.current;
+            slot.current = null;
+            if (current) {
+                current.reject(new WorkerTimeoutError(this.options.taskTimeoutMs));
+            }
+            // The worker is stuck in whatever wedged it, so it cannot be reused.
+            this.replace(slot);
+        }, this.options.taskTimeoutMs);
+        // A pending timer must not hold the process open on its own.
+        timer.unref?.();
+        slot.current = { id: item.envelope.id, resolve: item.resolve, reject: item.reject, timer };
+        slot.worker.postMessage(item.envelope);
+    }
+    drainQueue() {
+        while (this.queue.length) {
+            const slot = this.slots.find(candidate => !candidate.busy);
+            if (!slot) {
+                return;
+            }
+            const item = this.queue.shift();
+            if (!item) {
+                return;
+            }
+            this.dispatch(slot, item);
+        }
+    }
+    /** Number of tasks accepted but not yet completed. */
+    get inFlight() {
+        return this.queue.length + this.slots.filter(slot => slot.busy).length;
+    }
+    submit(task) {
+        if (this.closing) {
+            return Promise.reject(new Error('worker pool is closing'));
+        }
+        return new Promise((resolve, reject) => {
+            const item = { envelope: { id: this.nextId++, task }, resolve, reject };
+            const slot = this.slots.find(candidate => !candidate.busy);
+            if (slot) {
+                this.dispatch(slot, item);
+            }
+            else {
+                this.queue.push(item);
+            }
+        });
+    }
+    async close() {
+        this.closing = true;
+        while (this.queue.length) {
+            this.queue.shift()?.reject(new Error('worker pool is closing'));
+        }
+        const slots = [...this.slots];
+        this.slots.length = 0;
+        await Promise.all(slots.map(slot => {
+            if (slot.current) {
+                clearTimeout(slot.current.timer);
+            }
+            return slot.worker.terminate().catch(() => undefined);
+        }));
+    }
+}
+exports.WorkerPool = WorkerPool;
+//# sourceMappingURL=worker-pool.js.map

package/dist/util/worker-pool.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"worker-pool.js","sourceRoot":"","sources":["../../src/util/worker-pool.ts"],"names":[],"mappings":";AAAA;;;;;;;;GAQG;;;AAEH,6DAA6C;AAC7C,6CAA2C;AAiD3C,MAAa,kBAAmB,SAAQ,KAAK;IACzC,YAAY,EAAU;QAClB,KAAK,CAAC,+BAA+B,EAAE,IAAI,CAAC,CAAC;QAC7C,IAAI,CAAC,IAAI,GAAG,oBAAoB,CAAC;IACrC,CAAC;CACJ;AALD,gDAKC;AAED;;;GAGG;AACH,MAAa,UAA2B,SAAQ,0BAAY;IACvC,OAAO,CAAoB;IAC3B,KAAK,GAAoB,EAAE,CAAC;IAC5B,KAAK,GAA6B,EAAE,CAAC;IAC9C,MAAM,GAAG,CAAC,CAAC;IACX,OAAO,GAAG,KAAK,CAAC;IAExB,YAAY,OAA0B;QAClC,KAAK,EAAE,CAAC;QACR,IAAI,CAAC,OAAO,GAAG,OAAO,CAAC;QACvB,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,OAAO,CAAC,IAAI,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC;YACjD,IAAI,CAAC,KAAK,EAAE,CAAC;QACjB,CAAC;IACL,CAAC;IAEO,KAAK;QACT,MAAM,MAAM,GAAG,IAAI,4BAAM,CAAC,IAAI,CAAC,OAAO,CAAC,UAAU,EAAE;YAC/C,UAAU,EAAE,IAAI,CAAC,OAAO,CAAC,UAAU;YACnC,cAAc,EAAE,IAAI,CAAC,OAAO,CAAC,sBAAsB,CAAC,CAAC,CAAC,EAAE,sBAAsB,EAAE,IAAI,CAAC,OAAO,CAAC,sBAAsB,EAAE,CAAC,CAAC,CAAC,SAAS;SACpI,CAAC,CAAC;QACH,MAAM,IAAI,GAAkB,EAAE,MAAM,EAAE,IAAI,EAAE,KAAK,EAAE,OAAO,EAAE,IAAI,EAAE,CAAC;QAEnE,MAAM,CAAC,EAAE,CAAC,SAAS,EAAE,CAAC,KAAqB,EAAE,EAAE;YAC3C,MAAM,OAAO,GAAG,IAAI,CAAC,OAAO,CAAC;YAC7B,IAAI,CAAC,OAAO,IAAI,OAAO,CAAC,EAAE,KAAK,KAAK,CAAC,EAAE,EAAE,CAAC;gBACtC,OAAO;YACX,CAAC;YACD,YAAY,CAAC,OAAO,CAAC,KAAK,CAAC,CAAC;YAC5B,IAAI,CAAC,OAAO,GAAG,IAAI,CAAC;YACpB,IAAI,CAAC,IAAI,GAAG,KAAK,CAAC;YAClB,IAAI,KAAK,CAAC,EAAE,EAAE,CAAC;gBACX,OAAO,CAAC,OAAO,CAAC,KAAK,CAAC,MAAiB,CAAC,CAAC;YAC7C,CAAC;iBAAM,CAAC;gBACJ,OAAO,CAAC,MAAM,CAAC,IAAI,KAAK,CAAC,KAAK,CAAC,KAAK,IAAI,eAAe,CAAC,CAAC,CAAC;YAC9D,CAAC;YACD,IAAI,CAAC,UAAU,EAAE,CAAC;QACtB,CAAC,CAAC,CAAC;QAEH,MAAM,CAAC,EAAE,CAAC,OAAO,EAAE,GAAG,CAAC,EAAE;YACrB,IAAI,CAAC,IAAI,CAAC,cAAc,EAAE,GAAG,CAAC,CAAC;YAC/B,MAAM,OAAO,GAAG,IAAI,CAAC,OAAO,CAAC;YAC7B,IAAI,OAAO,EAAE,CAAC;gBACV,YAAY,CAAC,OAAO,CAAC,KAAK,CAAC,CAAC;gBAC5B,IAAI,CAAC,OAAO,GAAG,IAAI,CAAC;gBACpB,OAAO,CAAC,MAAM,CAAC,GAAG,YAAY,KAAK,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,IAAI,KAAK,CAAC,MAAM,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC;YACxE,CAAC;YACD,IAAI,CAAC,OAAO,CAAC,IAAI,CAAC,CAAC;QACvB,CAAC,CAAC,CAAC;QAEH,MAAM,CAAC,EAAE,CAAC,MAAM,EAAE,GAAG,EAAE;YACnB,IAAI,CAAC,IAAI,CAAC,OAAO,IAAI,IAAI,CAAC,KAAK,CAAC,QAAQ,CAAC,IAAI,CAAC,EAAE,CAAC;gBAC7C,IAAI,CAAC,OAAO,CAAC,IAAI,CAAC,CAAC;YACvB,CAAC;QACL,CAAC,CAAC,CAAC;QAEH,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QACtB,OAAO,IAAI,CAAC;IAChB,CAAC;IAEO,OAAO,CAAC,IAAmB;QAC/B,MAAM,KAAK,GAAG,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,IAAI,CAAC,CAAC;QACvC,IAAI,KAAK,KAAK,CAAC,CAAC,EAAE,CAAC;YACf,IAAI,CAAC,KAAK,CAAC,MAAM,CAAC,KAAK,EAAE,CAAC,CAAC,CAAC;QAChC,CAAC;QACD,IAAI,CAAC,IAAI,GAAG,KAAK,CAAC;QAClB,KAAK,IAAI,CAAC,MAAM,CAAC,SAAS,EAAE,CAAC,KAAK,CAAC,GAAG,EAAE,CAAC,SAAS,CAAC,CAAC;QACpD,IAAI,CAAC,IAAI,CAAC,OAAO,EAAE,CAAC;YAChB,IAAI,CAAC,KAAK,EAAE,CAAC;YACb,IAAI,CAAC,UAAU,EAAE,CAAC;QACtB,CAAC;IACL,CAAC;IAEO,QAAQ,CAAC,IAAmB,EAAE,IAA4B;QAC9D,IAAI,CAAC,IAAI,GAAG,IAAI,CAAC;QACjB,MAAM,KAAK,GAAG,UAAU,CAAC,GAAG,EAAE;YAC1B,MAAM,OAAO,GAAG,IAAI,CAAC,OAAO,CAAC;YAC7B,IAAI,CAAC,OAAO,GAAG,IAAI,CAAC;YACpB,IAAI,OAAO,EAAE,CAAC;gBACV,OAAO,CAAC,MAAM,CAAC,IAAI,kBAAkB,CAAC,IAAI,CAAC,OAAO,CAAC,aAAa,CAAC,CAAC,CAAC;YACvE,CAAC;YACD,qEAAqE;YACrE,IAAI,CAAC,OAAO,CAAC,IAAI,CAAC,CAAC;QACvB,CAAC,EAAE,IAAI,CAAC,OAAO,CAAC,aAAa,CAAC,CAAC;QAC/B,6DAA6D;QAC7D,KAAK,CAAC,KAAK,EAAE,EAAE,CAAC;QAEhB,IAAI,CAAC,OAAO,GAAG,EAAE,EAAE,EAAE,IAAI,CAAC,QAAQ,CAAC,EAAE,EAAE,OAAO,EAAE,IAAI,CAAC,OAAO,EAAE,MAAM,EAAE,IAAI,CAAC,MAAM,EAAE,KAAK,EAAE,CAAC;QAC3F,IAAI,CAAC,MAAM,CAAC,WAAW,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC;IAC3C,CAAC;IAEO,UAAU;QACd,OAAO,IAAI,CAAC,KAAK,CAAC,MAAM,EAAE,CAAC;YACvB,MAAM,IAAI,GAAG,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,SAAS,CAAC,EAAE,CAAC,CAAC,SAAS,CAAC,IAAI,CAAC,CAAC;YAC3D,IAAI,CAAC,IAAI,EAAE,CAAC;gBACR,OAAO;YACX,CAAC;YACD,MAAM,IAAI,GAAG,IAAI,CAAC,KAAK,CAAC,KAAK,EAAE,CAAC;YAChC,IAAI,CAAC,IAAI,EAAE,CAAC;gBACR,OAAO;YACX,CAAC;YACD,IAAI,CAAC,QAAQ,CAAC,IAAI,EAAE,IAAI,CAAC,CAAC;QAC9B,CAAC;IACL,CAAC;IAED,sDAAsD;IACtD,IAAI,QAAQ;QACR,OAAO,IAAI,CAAC,KAAK,CAAC,MAAM,GAAG,IAAI,CAAC,KAAK,CAAC,MAAM,CAAC,IAAI,CAAC,EAAE,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,MAAM,CAAC;IAC3E,CAAC;IAED,MAAM,CAAC,IAAW;QACd,IAAI,IAAI,CAAC,OAAO,EAAE,CAAC;YACf,OAAO,OAAO,CAAC,MAAM,CAAC,IAAI,KAAK,CAAC,wBAAwB,CAAC,CAAC,CAAC;QAC/D,CAAC;QACD,OAAO,IAAI,OAAO,CAAU,CAAC,OAAO,EAAE,MAAM,EAAE,EAAE;YAC5C,MAAM,IAAI,GAA2B,EAAE,QAAQ,EAAE,EAAE,EAAE,EAAE,IAAI,CAAC,MAAM,EAAE,EAAE,IAAI,EAAE,EAAE,OAAO,EAAE,MAAM,EAAE,CAAC;YAChG,MAAM,IAAI,GAAG,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,SAAS,CAAC,EAAE,CAAC,CAAC,SAAS,CAAC,IAAI,CAAC,CAAC;YAC3D,IAAI,IAAI,EAAE,CAAC;gBACP,IAAI,CAAC,QAAQ,CAAC,IAAI,EAAE,IAAI,CAAC,CAAC;YAC9B,CAAC;iBAAM,CAAC;gBACJ,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;YAC1B,CAAC;QACL,CAAC,CAAC,CAAC;IACP,CAAC;IAED,KAAK,CAAC,KAAK;QACP,IAAI,CAAC,OAAO,GAAG,IAAI,CAAC;QACpB,OAAO,IAAI,CAAC,KAAK,CAAC,MAAM,EAAE,CAAC;YACvB,IAAI,CAAC,KAAK,CAAC,KAAK,EAAE,EAAE,MAAM,CAAC,IAAI,KAAK,CAAC,wBAAwB,CAAC,CAAC,CAAC;QACpE,CAAC;QACD,MAAM,KAAK,GAAG,CAAC,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC;QAC9B,IAAI,CAAC,KAAK,CAAC,MAAM,GAAG,CAAC,CAAC;QACtB,MAAM,OAAO,CAAC,GAAG,CACb,KAAK,CAAC,GAAG,CAAC,IAAI,CAAC,EAAE;YACb,IAAI,IAAI,CAAC,OAAO,EAAE,CAAC;gBACf,YAAY,CAAC,IAAI,CAAC,OAAO,CAAC,KAAK,CAAC,CAAC;YACrC,CAAC;YACD,OAAO,IAAI,CAAC,MAAM,CAAC,SAAS,EAAE,CAAC,KAAK,CAAC,GAAG,EAAE,CAAC,SAAS,CAAC,CAAC;QAC1D,CAAC,CAAC,CACL,CAAC;IACN,CAAC;CACJ;AA5ID,gCA4IC"}

package/package.json

@@ -0,0 +1,74 @@
+{
+    "name": "@postalsys/email-content",
+    "version": "0.0.1",
+    "description": "Extract conversationally relevant content from arbitrary email messages",
+    "keywords": [
+        "email",
+        "mime",
+        "reply",
+        "quote",
+        "signature",
+        "extraction",
+        "parser"
+    ],
+    "license": "MIT",
+    "author": "Postal Systems OÜ",
+    "repository": {
+        "type": "git",
+        "url": "git+https://github.com/postalsys/email-content.git"
+    },
+    "bugs": {
+        "url": "https://github.com/postalsys/email-content/issues"
+    },
+    "homepage": "https://github.com/postalsys/email-content#readme",
+    "type": "commonjs",
+    "main": "dist/index.js",
+    "types": "dist/index.d.ts",
+    "exports": {
+        ".": {
+            "types": "./dist/index.d.ts",
+            "require": "./dist/index.js",
+            "default": "./dist/index.js"
+        },
+        "./package.json": "./package.json"
+    },
+    "bin": {
+        "email-content": "dist/cli/main.js"
+    },
+    "files": [
+        "dist"
+    ],
+    "engines": {
+        "node": ">=24.0.0"
+    },
+    "publishConfig": {
+        "access": "public"
+    },
+    "scripts": {
+        "build": "tsc -p tsconfig.json && node scripts/copy-assets.js",
+        "build:test": "tsc -p tsconfig.test.json && node scripts/copy-assets.js && node scripts/copy-assets-test.js",
+        "clean": "rm -rf dist .build-test",
+        "typecheck": "tsc -p tsconfig.test.json --noEmit",
+        "lint": "eslint src test",
+        "format": "prettier --write \"src/**/*.ts\" \"test/**/*.ts\" \"docs/**/*.md\" \"*.json\" \"*.js\"",
+        "format:check": "prettier --check \"src/**/*.ts\" \"test/**/*.ts\" \"docs/**/*.md\" \"*.json\" \"*.js\"",
+        "test": "npm run build:test && node --test \".build-test/test/**/*.test.js\"",
+        "test:fast": "node --test \"test/**/*.test.ts\"",
+        "prepublishOnly": "npm run build",
+        "cli": "node dist/cli/main.js"
+    },
+    "dependencies": {
+        "@zone-eu/mailsplit": "5.4.14",
+        "libmime": "5.4.1",
+        "node-html-parser": "9.0.0"
+    },
+    "devDependencies": {
+        "@eslint/js": "10.0.1",
+        "@types/node": "24.13.3",
+        "eslint": "10.8.0",
+        "eslint-config-prettier": "10.1.8",
+        "prettier": "3.9.6",
+        "typescript": "5.9.3",
+        "typescript-eslint": "8.65.0"
+    }
+}