Make tracking and publishing schedules configurable (#1086)

CHANGELOG.md

@@ -2,6 +2,16 @@
 
 All changes that impact users of this module are documented in this file, in the [Common Changelog](https://common-changelog.org) format with some additional specifications defined in the CONTRIBUTING file. This codebase adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## Unreleased [minor]
+
+> Development of this release was supported by the [French Ministry for Foreign Affairs](https://www.diplomatie.gouv.fr/fr/politique-etrangere-de-la-france/diplomatie-numerique/) through its ministerial [State Startups incubator](https://beta.gouv.fr/startups/open-terms-archive.html) under the aegis of the Ambassador for Digital Affairs.
+
+### Added
+
+- Make tracking schedule configurable
+- Make publishing schedule configurable
+- Add tracking overrun protection
+
 ## 2.0.1 - 2024-05-31
 
 _Full changeset and discussions: [#1085](https://github.com/OpenTermsArchive/engine/pull/1085)._

bin/ota-dataset.js

@@ -2,7 +2,9 @@
 import './env.js';
 
 import { program } from 'commander';
+import config from 'config';
 import cron from 'croner';
+import cronstrue from 'cronstrue';
 
 import { release } from '../scripts/dataset/index.js';
 import logger from '../src/logger/index.js';
@@ -26,8 +28,11 @@ const options = {
 if (!schedule) {
   await release(options);
 } else {
+  const trackingSchedule = config.get('@opentermsarchive/engine.dataset.publishingSchedule');
+  const humanReadableSchedule = cronstrue.toString(trackingSchedule);
+
   logger.info('The scheduler is running…');
-  logger.info('Dataset will be published every Monday at 08:30 in the timezone of this machine');
+  logger.info(`Dataset will be published ${humanReadableSchedule.toLowerCase()} in the timezone of this machine`);
 
-  cron('30 8 * * MON', () => release(options));
+  cron(config.get('@opentermsarchive/engine.dataset.publishingSchedule'), () => release(options));
 }

config/default.json

@@ -1,5 +1,6 @@
 {
   "@opentermsarchive/engine": {
+    "trackingSchedule": "30 */6 * * *",
     "services": {
       "declarationsPath": "./declarations"
     },
@@ -57,7 +58,8 @@
     },
     "dataset": {
       "title": "sandbox",
-      "versionsRepositoryURL": "https://github.com/OpenTermsArchive/sandbox"
+      "versionsRepositoryURL": "https://github.com/OpenTermsArchive/sandbox",
+      "publishingSchedule": "30 8 * * MON"
     }
   }
 }

package.json

@@ -62,7 +62,8 @@
     "chai-exclude": "^2.1.0",
     "commander": "^9.4.1",
     "config": "^3.3.6",
-    "croner": "^4.3.6",
+    "croner": "^8.0.2",
+    "cronstrue": "^2.50.0",
     "cross-env": "^7.0.3",
     "datauri": "^4.1.0",
     "dotenv": "^10.0.0",

src/index.js

@@ -1,5 +1,6 @@
 import config from 'config';
 import cron from 'croner';
+import cronstrue from 'cronstrue';
 
 import Archivist from './archivist/index.js';
 import logger from './logger/index.js';
@@ -71,8 +72,15 @@ export default async function track({ services, types, extractOnly, schedule })
     return;
   }
 
+  const trackingSchedule = config.get('@opentermsarchive/engine.trackingSchedule');
+  const humanReadableSchedule = cronstrue.toString(trackingSchedule);
+
   logger.info('The scheduler is running…');
-  logger.info('Terms will be tracked every six hours starting at half past midnight');
+  logger.info(`Terms will be tracked ${humanReadableSchedule.toLowerCase()} in the timezone of this machine`);
 
-  cron('30 */6 * * *', () => archivist.track({ services, types }));
+  cron(
+    trackingSchedule,
+    { protect: job => logger.warn(`Tracking scheduled at ${new Date().toISOString()} were blocked by an unfinished tracking started at ${job.currentRun().toISOString()}`) },
+    () => archivist.track({ services, types }),
+  );
 }