@crawlee/core 4.0.0-beta.98 → 4.0.0-rc.0

package/autoscaling/autoscaled_pool.d.ts

@@ -107,25 +107,7 @@ export interface AutoscaledPoolOptions extends TaskLoopPredicates {
  * @internal
  */
 export declare class AutoscaledPool {
-    private readonly log;
-    private readonly maybeRunIntervalMillis;
-    private readonly taskTimeoutMillis;
-    private readonly runTaskFunction;
-    private readonly isFinishedFunction;
-    private readonly isTaskReadyFunction;
-    private readonly concurrencySystem;
-    private readonly consumer;
-    private isStopped;
-    private resolve;
-    private reject;
-    private maybeRunInterval;
-    private queryingIsTaskReady;
-    private queryingIsFinished;
-    /**
-     * This pool's *own* in-flight task count, as opposed to {@link AutoscaledPool.currentConcurrency}, which is the
-     * (possibly shared) governor's total. `pause()` and `maybeFinish()` care only about this pool draining.
-     */
-    private ownConcurrency;
+    #private;
     constructor(options: AutoscaledPoolOptions);
     /**
      * The governor backing this pool, as supplied to the constructor — exposed as the read-only
@@ -194,13 +176,13 @@ export declare class AutoscaledPool {
      * Starts a new task
      * if the number of running tasks (current concurrency) is lower than desired concurrency
      * and the system is not currently overloaded
-     * and this.isTaskReadyFunction() returns true.
+     * and `isTaskReadyFunction()` returns true.
      *
      * It doesn't allow multiple concurrent runs of this method.
      */
     private maybeRunTask;
     /**
-     * If there are no running tasks and this.isFinishedFunction() returns true then closes
+     * If there are no running tasks and `isFinishedFunction()` returns true then closes
      * the pool and resolves the pool's promise returned by the run() method.
      *
      * It doesn't allow multiple concurrent runs of this method.

package/autoscaling/autoscaled_pool.js

@@ -1,8 +1,26 @@
-import ow from 'ow';
+import { z } from 'zod';
 import { addTimeoutToPromise } from '@apify/timeout';
 import { betterClearInterval, betterSetInterval } from '@apify/utilities';
 import { CriticalError } from '../errors.js';
 import { serviceLocator } from '../service_locator.js';
+import { parseArgument, schemas, validators } from '../validators.js';
+// `schemas.anyObject` and `objectWithKeys`-based validators pass values through by reference
+// (object schemas return a pruned plain copy), so class instances like loggers and the
+// concurrency system keep their prototype.
+const autoscaledPoolOptionsSchema = z.strictObject({
+    runTaskFunction: schemas.anyFunction,
+    isFinishedFunction: schemas.anyFunction,
+    isTaskReadyFunction: schemas.anyFunction,
+    maybeRunIntervalSecs: schemas.anyNumber
+        .refine((value) => value > 0, 'Expected a number greater than 0')
+        .default(0.5),
+    taskTimeoutSecs: schemas.anyNumber
+        .refine((value) => value >= 0, 'Expected a number greater than or equal to 0')
+        .default(0),
+    log: validators.logger.default(() => serviceLocator.getLogger()),
+    concurrencySystem: schemas.anyObject,
+    consumer: schemas.anyObject.refine((value) => typeof value.id === 'string' && value.id.length > 0, "Expected an object with a non-empty string 'id'"),
+});
 /**
  * Manages a pool of asynchronous resource-intensive tasks that are executed in parallel.
  * The pool only starts new tasks while its {@link IConcurrencySystem|concurrency system} reports free capacity —
@@ -56,52 +74,42 @@ import { serviceLocator } from '../service_locator.js';
  * @internal
  */
 export class AutoscaledPool {
-    log;
+    #log;
     // Configurable properties.
-    maybeRunIntervalMillis;
-    taskTimeoutMillis;
-    runTaskFunction;
-    isFinishedFunction;
-    isTaskReadyFunction;
-    concurrencySystem;
-    consumer;
+    #maybeRunIntervalMillis;
+    #taskTimeoutMillis;
+    #runTaskFunction;
+    #isFinishedFunction;
+    #isTaskReadyFunction;
+    #concurrencySystem;
+    #consumer;
     // Internal properties.
-    isStopped = false;
-    resolve = null;
-    reject = null;
-    maybeRunInterval;
-    queryingIsTaskReady;
-    queryingIsFinished;
+    #isStopped = false;
+    #resolve = null;
+    #reject = null;
+    #maybeRunInterval;
+    #queryingIsTaskReady;
+    #queryingIsFinished;
     /**
      * This pool's *own* in-flight task count, as opposed to {@link AutoscaledPool.currentConcurrency}, which is the
      * (possibly shared) governor's total. `pause()` and `maybeFinish()` care only about this pool draining.
      */
-    ownConcurrency = 0;
+    #ownConcurrency = 0;
     constructor(options) {
-        ow(options, ow.object.exactShape({
-            runTaskFunction: ow.function,
-            isFinishedFunction: ow.function,
-            isTaskReadyFunction: ow.function,
-            maybeRunIntervalSecs: ow.optional.number.greaterThan(0),
-            taskTimeoutSecs: ow.optional.number.greaterThanOrEqual(0),
-            log: ow.optional.object,
-            concurrencySystem: ow.object,
-            consumer: ow.object.partialShape({ id: ow.string.nonEmpty }),
-        }));
-        const { runTaskFunction, isFinishedFunction, isTaskReadyFunction, maybeRunIntervalSecs = 0.5, taskTimeoutSecs = 0, log = serviceLocator.getLogger(), concurrencySystem, consumer, } = options;
-        this.log = log.child({ prefix: 'AutoscaledPool' });
+        const { runTaskFunction, isFinishedFunction, isTaskReadyFunction, maybeRunIntervalSecs, taskTimeoutSecs, log, concurrencySystem, consumer, } = parseArgument(options, autoscaledPoolOptionsSchema, 'AutoscaledPoolOptions');
+        this.#log = log.child({ prefix: 'AutoscaledPool' });
         // Configurable properties.
-        this.maybeRunIntervalMillis = maybeRunIntervalSecs * 1000;
-        this.taskTimeoutMillis = taskTimeoutSecs * 1000;
-        this.runTaskFunction = runTaskFunction;
-        this.isFinishedFunction = isFinishedFunction;
-        this.isTaskReadyFunction = isTaskReadyFunction;
-        this.concurrencySystem = concurrencySystem;
-        this.consumer = consumer;
+        this.#maybeRunIntervalMillis = maybeRunIntervalSecs * 1000;
+        this.#taskTimeoutMillis = taskTimeoutSecs * 1000;
+        this.#runTaskFunction = runTaskFunction;
+        this.#isFinishedFunction = isFinishedFunction;
+        this.#isTaskReadyFunction = isTaskReadyFunction;
+        this.#concurrencySystem = concurrencySystem;
+        this.#consumer = consumer;
         // Internal properties.
-        this.isStopped = false;
-        this.resolve = null;
-        this.reject = null;
+        this.#isStopped = false;
+        this.#resolve = null;
+        this.#reject = null;
         this.maybeRunTask = this.maybeRunTask.bind(this);
     }
     /**
@@ -112,18 +120,18 @@ export class AutoscaledPool {
      * {@link ConcurrencySystem} its owner holds, never through the pool.
      */
     get system() {
-        return this.concurrencySystem;
+        return this.#concurrencySystem;
     }
     /** The estimated number of parallel tasks the governor can currently support. */
     get desiredConcurrency() {
-        return this.concurrencySystem.desiredConcurrency;
+        return this.#concurrencySystem.desiredConcurrency;
     }
     /**
      * The number of parallel tasks currently booked against the governor. When it is shared, this counts every
      * borrowing pool's tasks, not just this one's.
      */
     get currentConcurrency() {
-        return this.concurrencySystem.currentConcurrency;
+        return this.#concurrencySystem.currentConcurrency;
     }
     /**
      * Runs the auto-scaled pool. Returns a promise that gets resolved or rejected once
@@ -135,26 +143,26 @@ export class AutoscaledPool {
     async run() {
         // Checked here, on an awaited path — the capacity queries inside the task loop run from intervals and
         // `setImmediate`, where a throw would become an unhandled rejection and hang `run()` forever.
-        if (!this.concurrencySystem.isRunning) {
+        if (!this.#concurrencySystem.isRunning) {
             throw new CriticalError('The ConcurrencySystem this AutoscaledPool borrows has not been started, so system load would not be ' +
                 'monitored and the concurrency would never be adjusted. Whoever creates a ConcurrencySystem owns ' +
                 'its lifecycle: call `await concurrencySystem.start()` before running the pools or crawlers that ' +
                 'use it, and `await concurrencySystem.stop()` once they are all done.');
         }
         const poolPromise = new Promise((resolve, reject) => {
-            this.resolve = resolve;
-            this.reject = reject;
+            this.#resolve = resolve;
+            this.#reject = reject;
         });
         // This is here because if we scale down to let's say 1, then after each promise is finished
         // this.maybeRunTask() doesn't trigger another one. So if that 1 instance gets stuck it results
         // in the crawler getting stuck and even after scaling up it never triggers another promise.
-        this.maybeRunInterval = betterSetInterval(this.maybeRunTask, this.maybeRunIntervalMillis);
+        this.#maybeRunInterval = betterSetInterval(this.maybeRunTask, this.#maybeRunIntervalMillis);
         try {
             await poolPromise;
         }
         finally {
             // If resolve is null, the pool is already destroyed.
-            if (this.resolve)
+            if (this.#resolve)
                 await this.destroy();
         }
     }
@@ -170,9 +178,9 @@ export class AutoscaledPool {
      * parts of their asynchronous chains of commands will not execute.
      */
     async abort() {
-        this.isStopped = true;
-        if (this.resolve) {
-            this.resolve();
+        this.#isStopped = true;
+        if (this.#resolve) {
+            this.#resolve();
             await this.destroy();
         }
     }
@@ -192,9 +200,9 @@ export class AutoscaledPool {
      * it during a long pause, its owner can `stop()` and `start()` it again.
      */
     async pause(timeoutSecs) {
-        if (this.isStopped)
+        if (this.#isStopped)
             return;
-        this.isStopped = true;
+        this.#isStopped = true;
         await new Promise((resolve, reject) => {
             let timeout;
             let interval;
@@ -208,14 +216,14 @@ export class AutoscaledPool {
                 }, timeoutSecs);
             }
             interval = setInterval(() => {
-                if (this.ownConcurrency <= 0) {
+                if (this.#ownConcurrency <= 0) {
                     // Clean up timeout and interval to prevent process hanging.
                     if (timeout)
                         clearTimeout(timeout);
                     clearInterval(interval);
                     resolve();
                 }
-            }, this.maybeRunIntervalMillis);
+            }, this.#maybeRunIntervalMillis);
         });
     }
     /**
@@ -225,7 +233,7 @@ export class AutoscaledPool {
      * Tasks will automatically start running again in `options.maybeRunIntervalSecs`.
      */
     resume() {
-        this.isStopped = false;
+        this.#isStopped = false;
     }
     /**
      * Explicitly check the queue for new tasks. The AutoscaledPool checks the queue for new tasks periodically,
@@ -238,27 +246,27 @@ export class AutoscaledPool {
      * Starts a new task
      * if the number of running tasks (current concurrency) is lower than desired concurrency
      * and the system is not currently overloaded
-     * and this.isTaskReadyFunction() returns true.
+     * and `isTaskReadyFunction()` returns true.
      *
      * It doesn't allow multiple concurrent runs of this method.
      */
     async maybeRunTask(intervalCallback) {
-        this.log.perf('Attempting to run a task.');
+        this.#log.perf('Attempting to run a task.');
         // Check if the function was invoked by the maybeRunInterval and use an empty function if not.
         const done = intervalCallback || (() => { });
         // Prevent starting a new task if:
         // - the pool is paused or aborted
-        if (this.isStopped) {
-            this.log.perf('Task will not run. AutoscaledPool is stopped.');
+        if (this.#isStopped) {
+            this.#log.perf('Task will not run. AutoscaledPool is stopped.');
             return done();
         }
         // - we are already querying for a task.
-        if (this.queryingIsTaskReady) {
-            this.log.perf('Task will not run. Waiting for a ready task.');
+        if (this.#queryingIsTaskReady) {
+            this.#log.perf('Task will not run. Waiting for a ready task.');
             return done();
         }
         // - the budget has room for us.
-        if (!this.concurrencySystem.hasCapacityForTask(this.consumer)) {
+        if (!this.#concurrencySystem.hasCapacityForTask(this.#consumer)) {
             done();
             // A shared governor's budget can stay saturated by another pool indefinitely, so we still have to be able
             // to notice that *this* pool has run out of work — `maybeFinish()` is the only thing that ever resolves
@@ -267,37 +275,37 @@ export class AutoscaledPool {
             return this.maybeFinish();
         }
         // - a task is ready.
-        this.queryingIsTaskReady = true;
+        this.#queryingIsTaskReady = true;
         let isTaskReady;
         try {
-            this.log.perf('Checking for ready tasks.');
-            isTaskReady = await this.isTaskReadyFunction();
+            this.#log.perf('Checking for ready tasks.');
+            isTaskReady = await this.#isTaskReadyFunction();
         }
         catch (e) {
             const err = e;
-            this.log.perf('Checking for ready tasks failed.');
+            this.#log.perf('Checking for ready tasks failed.');
             // We might have already rejected this promise.
-            if (this.reject) {
+            if (this.#reject) {
                 // No need to log all concurrent errors.
-                this.log.exception(err, 'isTaskReadyFunction failed');
-                this.reject(err);
+                this.#log.exception(err, 'isTaskReadyFunction failed');
+                this.#reject(err);
             }
         }
         finally {
-            this.queryingIsTaskReady = false;
+            this.#queryingIsTaskReady = false;
         }
         if (!isTaskReady) {
-            this.log.perf('Task will not run. No tasks are ready.');
+            this.#log.perf('Task will not run. No tasks are ready.');
             done();
             // No tasks could mean that we're finished with all tasks.
             return this.maybeFinish();
         }
         // - the budget still has room. Re-checked atomically, because another pool sharing the governor may have taken
         // the last free slot while we awaited `isTaskReadyFunction` above.
-        if (!this.concurrencySystem.tryRegisterTaskStart(this.consumer)) {
+        if (!this.#concurrencySystem.tryRegisterTaskStart(this.#consumer)) {
             return done();
         }
-        this.ownConcurrency++;
+        this.#ownConcurrency++;
         try {
             // Everything's fine. Run task.
             // Try to run next task to build up concurrency,
@@ -306,73 +314,73 @@ export class AutoscaledPool {
             // We need to restart interval here, so that it doesn't get blocked by a stalled task.
             done();
             // Execute the current task.
-            this.log.perf('Running a task.');
-            if (this.taskTimeoutMillis > 0) {
-                await addTimeoutToPromise(async () => this.runTaskFunction(), this.taskTimeoutMillis, `runTaskFunction timed out after ${this.taskTimeoutMillis / 1000} seconds.`);
+            this.#log.perf('Running a task.');
+            if (this.#taskTimeoutMillis > 0) {
+                await addTimeoutToPromise(async () => this.#runTaskFunction(), this.#taskTimeoutMillis, `runTaskFunction timed out after ${this.#taskTimeoutMillis / 1000} seconds.`);
             }
             else {
-                await this.runTaskFunction();
+                await this.#runTaskFunction();
             }
-            this.log.perf('Task finished.');
+            this.#log.perf('Task finished.');
             // Run task after the previous one finished. Only on success: a failed task rejects the pool, and
             // nudging the loop afterwards could start work on an already destroyed pool.
             setImmediate(this.maybeRunTask);
         }
         catch (e) {
             const err = e;
-            this.log.perf('Running a task failed.');
+            this.#log.perf('Running a task failed.');
             // We might have already rejected this promise.
-            if (this.reject) {
+            if (this.#reject) {
                 // No need to log all concurrent errors.
                 if (
                 // avoid reprinting the same critical error multiple times, as it will be printed by Nodejs at the end anyway
                 !(e instanceof CriticalError)) {
-                    this.log.exception(err, 'runTaskFunction failed.');
+                    this.#log.exception(err, 'runTaskFunction failed.');
                 }
-                this.reject(err);
+                this.#reject(err);
             }
         }
         finally {
-            this.concurrencySystem.registerTaskEnd(this.consumer);
-            this.ownConcurrency--;
+            this.#concurrencySystem.registerTaskEnd(this.#consumer);
+            this.#ownConcurrency--;
         }
         return undefined;
     }
     /**
-     * If there are no running tasks and this.isFinishedFunction() returns true then closes
+     * If there are no running tasks and `isFinishedFunction()` returns true then closes
      * the pool and resolves the pool's promise returned by the run() method.
      *
      * It doesn't allow multiple concurrent runs of this method.
      */
     async maybeFinish() {
-        if (this.queryingIsFinished)
+        if (this.#queryingIsFinished)
             return;
-        if (this.ownConcurrency > 0)
+        if (this.#ownConcurrency > 0)
             return;
-        this.queryingIsFinished = true;
+        this.#queryingIsFinished = true;
         try {
-            const isFinished = await this.isFinishedFunction();
-            if (isFinished && this.resolve)
-                this.resolve();
+            const isFinished = await this.#isFinishedFunction();
+            if (isFinished && this.#resolve)
+                this.#resolve();
         }
         catch (e) {
             const err = e;
-            if (this.reject) {
+            if (this.#reject) {
                 // No need to log all concurrent errors.
-                this.log.exception(err, 'isFinishedFunction failed.');
-                this.reject(err);
+                this.#log.exception(err, 'isFinishedFunction failed.');
+                this.#reject(err);
             }
         }
         finally {
-            this.queryingIsFinished = false;
+            this.#queryingIsFinished = false;
         }
     }
     /**
      * Cleans up resources.
      */
     async destroy() {
-        this.resolve = null;
-        this.reject = null;
-        betterClearInterval(this.maybeRunInterval);
+        this.#resolve = null;
+        this.#reject = null;
+        betterClearInterval(this.#maybeRunInterval);
     }
 }

package/autoscaling/concurrency_system.d.ts

@@ -53,8 +53,8 @@ export interface ConcurrencySystemOptions {
     autoscaleIntervalSecs?: number;
     /**
      * The signals that tell the system whether the machine is overloaded: per-resource tuning for the built-in four
-     * (memory, event loop, CPU, client) plus any {@link LoadSignalsOptions.custom|`custom`} implementations of
-     * your own. See {@link LoadSignalsOptions}.
+     * (memory, event loop, CPU, storage backend) plus any {@link LoadSignalsOptions.custom|`custom`}
+     * implementations of your own. See {@link LoadSignalsOptions}.
      */
     loadSignals?: LoadSignalsOptions;
     /**
@@ -152,30 +152,15 @@ export interface IConcurrencySystem {
  * @category Scaling
  */
 export declare class ConcurrencySystem implements IConcurrencySystem {
+    #private;
     private readonly log;
     private readonly desiredConcurrencyRatio;
     private readonly scaleUpStepRatio;
     private readonly scaleDownStepRatio;
-    private readonly loggingIntervalMillis;
-    private readonly autoscaleIntervalMillis;
     private readonly maxTasksPerMinute;
-    private _minConcurrency;
-    private _maxConcurrency;
-    private _desiredConcurrency;
     private _currentConcurrency;
-    private lastLoggingTime?;
-    private _tasksPerMinute;
     private readonly snapshotter;
-    private readonly loadSignals;
     private readonly systemStatus;
-    private autoscaleInterval?;
-    private tasksDonePerSecondInterval?;
-    /** Whether the snapshotter and autoscaling intervals are currently running. */
-    private running;
-    /** The in-flight (or completed) startup, memoized so concurrent `start()` calls await one boot. */
-    private startPromise?;
-    /** Set once per session, so a pool outliving `stop()` is reported once rather than every half second. */
-    private warnedAboutQueryWhileStopped;
     constructor(options?: ConcurrencySystemOptions);
     /**
      * Gets the minimum number of tasks running in parallel.
@@ -270,14 +255,14 @@ export declare class ConcurrencySystem implements IConcurrencySystem {
      * Evaluates the historical system status and scales the shared desired concurrency up or down accordingly. Driven
      * by the autoscaling interval started in {@link ConcurrencySystem.start|`start()`}.
      */
-    private _autoscale;
+    private autoscale;
     /**
      * Scales the system up by increasing the desired concurrency by the scaleUpStepRatio.
      */
-    private _scaleUp;
+    private scaleUp;
     /**
      * Scales the system down by decreasing the desired concurrency by the scaleDownStepRatio.
      */
-    private _scaleDown;
-    private _incrementTasksDonePerSecond;
+    private scaleDown;
+    private incrementTasksDonePerSecond;
 }