Press n or j to go to the next uncovered block, b, p or k for the previous block.
| 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 | 4x 4x 4x 344x 441x 58x 69x 69x 69x 69x 69x 69x 69x 3x 3x 3x 69x 69x 69x 69x 9x 9x 9x 9x 11x 11x 11x 9x 18x 1x 17x 17x 17x 24x 24x 38x 24x 34x 24x 17x 24x 7x 18x 11x 11x 10x 16x 16x 8x 8x 8x 8x 8x 8x 24x 22x 17x 24x 24x 75x 64x 24x 16x 16x 16x 16x 8x 8x 8x 24x 24x 32x 16x 24x 11x 8x 24x 5x 5x 1x 4x 4x 4x 17x 17x 6x 6x 6x 6x 10x 10x 10x 6x 13x 1x 12x 12x 12x 12x 1x 11x 11x 11x 11x 11x 11x 33x 18x 18x 28x 15x 15x 21x 17x 17x 17x 15x 11x 11x 11x 11x 6x 5x 11x 11x 1x 11x 16x 3x 13x 16x 9x 9x 9x 7x 7x 7x 2x 2x 16x 4x 4x 4x 4x 13x 5x 5x 13x 16x 16x 16x 13x 13x 13x 9x 12x 12x 12x 12x 12x 12x 270x 311x 304x 311x 301x 28x 28x 26x 80x 4x 11x 11x 11x 11x 11x 11x 11x 11x | // SPDX-FileCopyrightText: 2024-2026 Hack23 AB
// SPDX-License-Identifier: Apache-2.0
/**
* @module Utils/IntelligenceIndex
* @description Cross-article intelligence indexing system for parliamentary trend tracking.
*
* Provides functions to build, query, and persist an {@link IntelligenceIndex} that
* tracks relationships between generated articles, detects recurring trends, and groups
* related articles into series. The index is JSON-serialisable for disk persistence.
*/
import fs from 'fs';
import path from 'path';
import type {
ArticleCrossReference,
ArticleIndexEntry,
ArticleSeries,
IntelligenceIndex,
TrendDetection,
} from '../types/index.js';
import { ArticleCategory } from '../types/index.js';
import type { RelationshipLabels } from '../constants/languages.js';
import { RELATED_ANALYSIS_LABELS, getLocalizedString } from '../constants/languages.js';
// ─── Minimum article count to confirm a trend ────────────────────────────────
/** Minimum number of articles required to recognise a trend */
const MIN_TREND_ARTICLES = 2;
/** Maximum cross-reference results when no limit is specified */
const DEFAULT_MAX_RELATED = 10;
/**
* Keys that must never be used as lookup-map indices to prevent
* prototype-pollution attacks from untrusted article metadata.
*
* - `__proto__` — directly sets the prototype chain of plain objects
* - `constructor` — can be used to reach `Object` and mutate shared prototypes
* - `prototype` — when combined with `constructor`, enables prototype injection
*/
const DANGEROUS_KEYS = new Set(['__proto__', 'constructor', 'prototype']);
/**
* Return `true` when `key` is safe to use as a plain-object property.
* Rejects `__proto__`, `constructor`, and `prototype` to avoid prototype
* pollution when indexing untrusted topic/actor/procedure strings.
*
* @param key - The key to validate
* @returns `true` if the key is safe to use as an object property
*/
function isSafeKey(key: string): boolean {
return !DANGEROUS_KEYS.has(key);
}
/**
* Create a null-prototype object suitable for use as a lookup map.
* Unlike `{}`, these objects have no inherited keys, providing an extra
* layer of defence against prototype-pollution.
*
* @returns A fresh null-prototype Record for use as a lookup map
*/
function createNullMap(): Record<string, string[]> {
return Object.create(null) as Record<string, string[]>;
}
// ─── createEmptyIndex ────────────────────────────────────────────────────────
/**
* Create a fresh, empty {@link IntelligenceIndex}.
*
* @returns An empty index with no articles, actors, domains, trends, or series
*/
export function createEmptyIndex(): IntelligenceIndex {
return {
articles: [],
actors: createNullMap(),
policyDomains: createNullMap(),
procedures: createNullMap(),
trends: [],
series: [],
lastUpdated: new Date().toISOString(),
};
}
// ─── addArticleToIndex ───────────────────────────────────────────────────────
/**
* Add an {@link ArticleIndexEntry} to the index and keep all lookup maps in sync.
*
* The function is immutable-safe: it returns a new index object rather than
* mutating the supplied one. Duplicate entries (same `id`) are replaced.
*
* @param index - Existing intelligence index
* @param entry - Article index entry to add
* @returns Updated index with the new entry reflected in all maps
*/
export function addArticleToIndex(
index: IntelligenceIndex,
entry: ArticleIndexEntry
): IntelligenceIndex {
// Replace or append
const existingIdx = index.articles.findIndex((a) => a.id === entry.id);
const oldEntry = existingIdx >= 0 ? index.articles[existingIdx] : undefined; // eslint-disable-line security/detect-object-injection -- existingIdx from findIndex
const articles =
existingIdx >= 0
? [...index.articles.slice(0, existingIdx), entry, ...index.articles.slice(existingIdx + 1)]
: [...index.articles, entry];
// Clone lookup maps (null-prototype to prevent pollution)
const actors = Object.assign(createNullMap(), index.actors);
const policyDomains = Object.assign(createNullMap(), index.policyDomains);
const procedures = Object.assign(createNullMap(), index.procedures);
// Remove stale associations from the old entry (if replacing)
if (oldEntry) {
removeIdFromMap(actors, oldEntry.keyActors, entry.id);
removeIdFromMap(policyDomains, oldEntry.keyTopics, entry.id);
removeIdFromMap(procedures, oldEntry.procedures, entry.id);
}
// Add new associations
addIdToMap(actors, entry.keyActors, entry.id);
addIdToMap(policyDomains, entry.keyTopics, entry.id);
addIdToMap(procedures, entry.procedures, entry.id);
return {
...index,
articles,
actors,
policyDomains,
procedures,
lastUpdated: new Date().toISOString(),
};
}
// ─── buildIndexFromEntries ────────────────────────────────────────────────────
/**
* Build a complete index from an array of entries in O(n) time.
*
* Unlike calling {@link addArticleToIndex} in a loop (which clones maps on every
* call, yielding O(n²) behaviour), this function mutates local maps in a single
* pass and returns a final immutable index.
*
* @param entries - All article entries to include
* @returns A fully populated {@link IntelligenceIndex}
*/
export function buildIndexFromEntries(entries: ArticleIndexEntry[]): IntelligenceIndex {
const actors = createNullMap();
const policyDomains = createNullMap();
const procedures = createNullMap();
for (const entry of entries) {
addIdToMap(actors, entry.keyActors, entry.id);
addIdToMap(policyDomains, entry.keyTopics, entry.id);
addIdToMap(procedures, entry.procedures, entry.id);
}
return {
articles: [...entries],
actors,
policyDomains,
procedures,
trends: [],
series: [],
lastUpdated: new Date().toISOString(),
};
}
// ─── findRelatedArticles ─────────────────────────────────────────────────────
/**
* Find articles that share topics or actors with the supplied lists.
*
* Results are scored by overlap count and returned in descending relevance order.
*
* @param index - Intelligence index to search
* @param topics - Key topics to match against `keyTopics`
* @param actors - Key actors to match against `keyActors`
* @param maxResults - Maximum number of results to return (default: 10)
* @returns Scored, sorted array of matching article entries
*/
export function findRelatedArticles(
index: IntelligenceIndex,
topics: string[],
actors: string[],
maxResults: number = DEFAULT_MAX_RELATED
): ArticleIndexEntry[] {
if (topics.length === 0 && actors.length === 0) {
return [];
}
const topicSet = new Set(topics);
const actorSet = new Set(actors);
const scored = index.articles.map((article) => {
let score = 0;
for (const t of article.keyTopics) {
if (topicSet.has(t)) score++;
}
for (const a of article.keyActors) {
if (actorSet.has(a)) score++;
}
return { article, score };
});
return scored
.filter(({ score }) => score > 0)
.sort((a, b) => b.score - a.score)
.slice(0, maxResults)
.map(({ article }) => article);
}
// ─── generateCrossReferences ─────────────────────────────────────────────────
/**
* Auto-generate {@link ArticleCrossReference} objects for an article based on
* topic and actor overlap with existing index entries.
*
* **Strength** is determined by total overlap (topics + actors):
* - ≥3 shared items → `strong`
* - 2 shared items → `moderate`
* - 1 shared item → `weak`
*
* **Relationship** is determined by date comparison:
* - Target article is older (`date < entry.date`) → `follows_up`
* - Target article is newer (`date > entry.date`) → `preceded_by`
* - Target article has the same date → `related`
*
* @param index - Intelligence index containing previously indexed articles
* @param entry - The article for which cross-references should be generated
* @returns Array of auto-generated cross-references (excludes self-references)
*/
export function generateCrossReferences(
index: IntelligenceIndex,
entry: ArticleIndexEntry
): ArticleCrossReference[] {
const related = findRelatedArticles(index, entry.keyTopics, entry.keyActors);
return related
.filter((a) => a.id !== entry.id)
.map((a) => {
const topicOverlap = a.keyTopics.filter((t) => entry.keyTopics.includes(t)).length;
const actorOverlap = a.keyActors.filter((ac) => entry.keyActors.includes(ac)).length;
const totalOverlap = topicOverlap + actorOverlap;
const strength: ArticleCrossReference['strength'] =
totalOverlap >= 3 ? 'strong' : totalOverlap === 2 ? 'moderate' : 'weak';
const relationship: ArticleCrossReference['relationship'] =
a.date < entry.date ? 'follows_up' : a.date > entry.date ? 'preceded_by' : 'related';
const context =
topicOverlap > 0 && actorOverlap > 0
? `Shares ${topicOverlap} topic(s) and ${actorOverlap} actor(s)`
: topicOverlap > 0
? `Shares ${topicOverlap} topic(s)`
: `Shares ${actorOverlap} actor(s)`;
const ref: ArticleCrossReference = {
targetArticleId: a.id,
relationship,
context,
strength,
};
return ref;
});
}
// ─── detectTrends ────────────────────────────────────────────────────────────
/**
* Resolve confidence level from the number of article references.
* @param count - Number of articles referencing the topic
* @returns Confidence level
*/
function resolveConfidence(count: number): TrendDetection['confidence'] {
if (count >= 5) return 'high';
if (count >= 3) return 'medium';
return 'low';
}
/**
* Resolve date range from a list of articles matching a set of IDs.
* @param index - The intelligence index
* @param articleIds - IDs of articles to consider
* @returns firstSeen and lastUpdated date strings
*/
function resolveDateRange(
index: IntelligenceIndex,
articleIds: string[]
): { firstSeen: string; lastUpdated: string } {
const fallback = new Date().toISOString().slice(0, 10);
const dates = index.articles
.filter((a) => articleIds.includes(a.id))
.map((a) => a.date)
.sort();
return {
firstSeen: dates[0] ?? fallback,
lastUpdated: dates[dates.length - 1] ?? fallback,
};
}
/**
* Build a topic-based trend entry.
* @param index - Intelligence index
* @param topic - Topic key
* @param articleIds - Article IDs covering this topic
* @returns TrendDetection entry
*/
function buildTopicTrend(
index: IntelligenceIndex,
topic: string,
articleIds: string[]
): TrendDetection {
const { firstSeen, lastUpdated } = resolveDateRange(index, articleIds);
const confidence = resolveConfidence(articleIds.length);
const direction: TrendDetection['direction'] =
articleIds.length >= 4 ? 'strengthening' : 'emerging';
return {
id: `trend-topic-${slugify(topic)}`,
name: `${topic} trend`,
category: 'political',
direction,
firstSeen,
lastUpdated,
articleReferences: [...articleIds],
evidence: [`Topic "${topic}" covered in ${articleIds.length} articles`],
confidence,
};
}
/**
* Build a procedure-based trend entry.
* @param index - Intelligence index
* @param proc - Procedure reference
* @param articleIds - Article IDs covering this procedure
* @returns TrendDetection entry
*/
function buildProcedureTrend(
index: IntelligenceIndex,
proc: string,
articleIds: string[]
): TrendDetection {
const { firstSeen, lastUpdated } = resolveDateRange(index, articleIds);
const confidence = resolveConfidence(articleIds.length);
return {
id: `trend-proc-${slugify(proc)}`,
name: `Procedure ${proc} tracking`,
category: 'legislative',
direction: 'stable',
firstSeen,
lastUpdated,
articleReferences: [...articleIds],
evidence: [`Procedure "${proc}" tracked across ${articleIds.length} articles`],
confidence,
};
}
/**
* Detect parliamentary trends from patterns across all indexed articles.
*
* A trend is formed when a topic or procedure appears in at least
* {@link MIN_TREND_ARTICLES} articles. The returned array replaces any
* previously detected trends stored in the index.
*
* @param index - Intelligence index to analyse
* @returns Array of detected {@link TrendDetection} objects
*/
export function detectTrends(index: IntelligenceIndex): TrendDetection[] {
const trends: TrendDetection[] = [];
for (const [topic, articleIds] of Object.entries(index.policyDomains)) {
if (articleIds.length >= MIN_TREND_ARTICLES) {
trends.push(buildTopicTrend(index, topic, articleIds));
}
}
for (const [proc, articleIds] of Object.entries(index.procedures)) {
if (articleIds.length >= MIN_TREND_ARTICLES) {
trends.push(buildProcedureTrend(index, proc, articleIds));
}
}
return trends;
}
// ─── findOrCreateSeries ──────────────────────────────────────────────────────
/**
* Find an existing {@link ArticleSeries} for the given procedure reference, or
* create a new one and add it to the index.
*
* **Note:** This function mutates `index.series` for convenience. Callers that
* require immutability should replace the series array after calling this.
*
* @param index - Intelligence index to search / mutate
* @param procedureRef - EP procedure reference (e.g. "2024/0001(COD)")
* @param name - Display name for the series if it needs to be created
* @returns The found or newly created series
*/
export function findOrCreateSeries(
index: IntelligenceIndex,
procedureRef: string,
name: string
): ArticleSeries {
const existing = index.series.find((s) => s.procedureRef === procedureRef);
if (existing) {
return existing;
}
const newSeries: ArticleSeries = {
id: `series-${slugify(procedureRef)}`,
name,
procedureRef,
articles: [],
status: 'ongoing',
summary: `Tracking legislative procedure ${procedureRef}`,
};
index.series.push(newSeries);
return newSeries;
}
// ─── normalizeArticleEntry ────────────────────────────────────────────────────
/**
* Ensure an {@link ArticleIndexEntry} loaded from disk has all required fields.
*
* Persisted JSON may be missing arrays or strings after schema evolution; this
* function fills safe defaults so downstream code never crashes on `undefined`.
*
* @param entry - Potentially partial article entry from parsed JSON
* @returns A fully populated {@link ArticleIndexEntry}
*/
function normalizeArticleEntry(entry: Partial<ArticleIndexEntry>): ArticleIndexEntry {
const VALID_CATEGORIES = new Set<string>(Object.values(ArticleCategory));
return {
id: typeof entry.id === 'string' ? entry.id : '',
date: typeof entry.date === 'string' ? entry.date : '',
type:
typeof entry.type === 'string' && VALID_CATEGORIES.has(entry.type)
? entry.type
: (ArticleCategory.WEEK_AHEAD as ArticleIndexEntry['type']),
lang: typeof entry.lang === 'string' ? entry.lang : 'en',
keyTopics: Array.isArray(entry.keyTopics) ? entry.keyTopics : [],
keyActors: Array.isArray(entry.keyActors) ? entry.keyActors : [],
procedures: Array.isArray(entry.procedures) ? entry.procedures : [],
crossReferences: Array.isArray(entry.crossReferences) ? entry.crossReferences : [],
trendContributions: Array.isArray(entry.trendContributions) ? entry.trendContributions : [],
...(typeof entry.seriesId === 'string' && { seriesId: entry.seriesId }),
};
}
// ─── rebuildLookupMaps ──────────────────────────────────────────────────────
/**
* Rebuild the actor/policyDomain/procedure reverse-lookup maps from scratch
* by scanning all articles.
*
* Used during index loading when persisted maps are missing or invalid (e.g.
* after a schema evolution), so that trend detection and lookups stay correct.
*
* @param articles - Normalised article entries to rebuild maps from
* @returns Rebuilt lookup maps
*/
function rebuildLookupMaps(articles: readonly ArticleIndexEntry[]): {
actors: Record<string, string[]>;
policyDomains: Record<string, string[]>;
procedures: Record<string, string[]>;
} {
const actors = createNullMap();
const policyDomains = createNullMap();
const procedures = createNullMap();
for (const article of articles) {
addIdToMap(actors, article.keyActors, article.id);
addIdToMap(policyDomains, article.keyTopics, article.id);
addIdToMap(procedures, article.procedures, article.id);
}
return { actors, policyDomains, procedures };
}
// ─── loadIntelligenceIndex ───────────────────────────────────────────────────
/**
* Load an {@link IntelligenceIndex} from a JSON file.
*
* Returns an empty index (via {@link createEmptyIndex}) if the file does not exist
* or cannot be parsed.
*
* Each loaded {@link ArticleIndexEntry} is normalised so that missing arrays
* and strings are filled with safe defaults, preventing crashes in downstream
* code that iterates over `keyTopics`, `keyActors`, etc.
*
* When any lookup map (actors, policyDomains, procedures) is missing or invalid
* but articles are present, all maps are rebuilt from the article entries so that
* trend detection and lookups remain correct across schema upgrades.
*
* @param indexPath - Absolute or relative path to the index JSON file
* @returns Loaded index, or a fresh empty index on failure
*/
export function loadIntelligenceIndex(indexPath: string): IntelligenceIndex {
if (!fs.existsSync(indexPath)) {
return createEmptyIndex();
}
try {
const content = fs.readFileSync(indexPath, 'utf-8');
const parsed = JSON.parse(content) as Partial<IntelligenceIndex>;
return mergeOntoEmpty(parsed);
} catch {
return createEmptyIndex();
}
}
/**
* Merge partially-parsed JSON onto a fresh empty index, normalising articles and
* rebuilding lookup maps when they are missing or invalid. When maps are rebuilt
* (or when persisted trends are missing), trends are also recomputed from articles
* so trend detection stays correct across schema upgrades.
*
* @param parsed - Potentially partial index from disk
* @returns Fully populated {@link IntelligenceIndex}
*/
function mergeOntoEmpty(parsed: Partial<IntelligenceIndex>): IntelligenceIndex {
const empty = createEmptyIndex();
const articles = Array.isArray(parsed.articles)
? parsed.articles.map(normalizeArticleEntry)
: empty.articles;
const { actors, policyDomains, procedures, rebuilt } = resolveOrRebuildMaps(
parsed,
articles,
empty
);
// Build a temporary index so detectTrends can recompute from articles
const base: IntelligenceIndex = {
articles,
actors,
policyDomains,
procedures,
trends: empty.trends,
series: Array.isArray(parsed.series) ? parsed.series : empty.series,
lastUpdated: typeof parsed.lastUpdated === 'string' ? parsed.lastUpdated : empty.lastUpdated,
};
// Recompute trends when maps were rebuilt or when persisted trends are missing/invalid
const trends = rebuilt || !Array.isArray(parsed.trends) ? detectTrends(base) : parsed.trends;
return { ...base, trends };
}
/**
* Check whether a value is a valid lookup map (`Record<string, string[]>`).
*
* Validates that the value is a non-array object **and** that every entry
* is an array of strings. Malformed/corrupt JSON (e.g. `{ EPP: "a1" }`)
* will return `false`, triggering a map rebuild from articles on load.
*
* @param value - Value to validate
* @returns `true` if the value is a well-formed lookup map
*/
function isValidMap(value: unknown): value is Record<string, string[]> {
if (!value || typeof value !== 'object' || Array.isArray(value)) return false;
const record = value as Record<string, unknown>;
return Object.values(record).every(
(v) => Array.isArray(v) && v.every((item) => typeof item === 'string')
);
}
/**
* Copy a structurally valid parsed map into a fresh null-prototype object,
* filtering out dangerous keys to maintain prototype-pollution protection
* even when loading from potentially tampered JSON.
*
* @param source - Validated lookup map from parsed JSON
* @returns A sanitised null-prototype copy with dangerous keys removed
*/
function sanitizeMap(source: Record<string, string[]>): Record<string, string[]> {
const safe = createNullMap();
for (const key of Object.keys(source)) {
if (!isSafeKey(key)) continue;
// eslint-disable-next-line security/detect-object-injection -- key validated via isSafeKey
const val = source[key];
Eif (val) {
// eslint-disable-next-line security/detect-object-injection -- key validated via isSafeKey
safe[key] = val;
}
}
return safe;
}
/**
* Return lookup maps from the parsed JSON when all three are valid, or rebuild
* them from the article entries when any map is missing/invalid.
*
* Valid parsed maps are always copied into fresh null-prototype objects with
* unsafe keys filtered out, so that tampered index JSON cannot reintroduce
* prototype-pollution vectors (e.g. `__proto__` keys).
*
* @param parsed - Partially parsed index from disk
* @param articles - Normalised article entries
* @param empty - Fallback empty index for defaults
* @returns Resolved or rebuilt lookup maps
*/
function resolveOrRebuildMaps(
parsed: Partial<IntelligenceIndex>,
articles: ArticleIndexEntry[],
empty: IntelligenceIndex
): {
actors: Record<string, string[]>;
policyDomains: Record<string, string[]>;
procedures: Record<string, string[]>;
rebuilt: boolean;
} {
const validActors = isValidMap(parsed.actors);
const validDomains = isValidMap(parsed.policyDomains);
const validProcedures = isValidMap(parsed.procedures);
if (articles.length > 0 && (!validActors || !validDomains || !validProcedures)) {
return { ...rebuildLookupMaps(articles), rebuilt: true };
}
return {
actors: validActors ? sanitizeMap(parsed.actors as Record<string, string[]>) : empty.actors,
policyDomains: validDomains
? sanitizeMap(parsed.policyDomains as Record<string, string[]>)
: empty.policyDomains,
procedures: validProcedures
? sanitizeMap(parsed.procedures as Record<string, string[]>)
: empty.procedures,
rebuilt: false,
};
}
// ─── saveIntelligenceIndex ───────────────────────────────────────────────────
/**
* Persist an {@link IntelligenceIndex} to a JSON file.
*
* Creates any missing parent directories automatically.
*
* @param index - Intelligence index to save
* @param indexPath - Absolute or relative path to the output JSON file
*/
export function saveIntelligenceIndex(index: IntelligenceIndex, indexPath: string): void {
const dir = path.dirname(indexPath);
if (!fs.existsSync(dir)) {
fs.mkdirSync(dir, { recursive: true });
}
fs.writeFileSync(indexPath, JSON.stringify(index, null, 2), 'utf-8');
}
// ─── buildRelatedArticlesHTML ────────────────────────────────────────────────
/**
* Generate an HTML `<section>` listing related articles, cross-references, and
* emerging trends for embedding in a generated article.
*
* Produces accessible markup with `aria-label` and `rel="noopener noreferrer"`.
* UI strings and date formatting are localised based on the `lang` parameter.
*
* @param relatedArticles - Articles related to the current article
* @param crossRefs - Cross-references from the current article
* @param trends - Trends relevant to the current article
* @param lang - Language code for localisation (defaults to 'en')
* @returns HTML string for the "Related Analysis" section, or empty string if nothing to show
*/
export function buildRelatedArticlesHTML(
relatedArticles: ArticleIndexEntry[],
crossRefs: ArticleCrossReference[],
trends: TrendDetection[],
lang?: string
): string {
if (relatedArticles.length === 0 && crossRefs.length === 0 && trends.length === 0) {
return '';
}
const strings = getLocalizedString(RELATED_ANALYSIS_LABELS, lang ?? 'en');
const listItems = crossRefs
.map((ref) => {
const article = relatedArticles.find((a) => a.id === ref.targetArticleId);
const label =
strings.relationships[ref.relationship as keyof RelationshipLabels] ??
strings.relatedArticle;
if (article) {
const displayDate = formatDisplayDate(article.date, lang);
const filename = `${article.id}.html`;
return ` <li><a href="${escapeAttr(filename)}" rel="noopener noreferrer">${escapeText(label)}: ${escapeText(ref.context)} (${escapeText(displayDate)})</a></li>`;
}
// Fallback: render using targetArticleId when full article metadata is unavailable
const filename = `${ref.targetArticleId}.html`;
return ` <li><a href="${escapeAttr(filename)}" rel="noopener noreferrer">${escapeText(label)}: ${escapeText(ref.context)}</a></li>`;
})
.filter(Boolean);
// Fall back: show related articles without explicit cross-refs
if (listItems.length === 0 && relatedArticles.length > 0) {
for (const article of relatedArticles) {
const displayDate = formatDisplayDate(article.date, lang);
const filename = `${article.id}.html`;
listItems.push(
` <li><a href="${escapeAttr(filename)}" rel="noopener noreferrer">${escapeText(strings.relatedArticle)}: ${escapeText(article.type)} (${escapeText(displayDate)})</a></li>`
);
}
}
const trendBlocks = trends
.map((trend) => {
const count = trend.articleReferences.length;
return ` <div class="emerging-trends">
<h4>${escapeText(strings.emergingTrend)}: ${escapeText(trend.name)}</h4>
<p>${count} ${escapeText(strings.trendTracking)} ${escapeText(trend.name.toLowerCase())} (${escapeText(strings.confidence)}: ${escapeText(trend.confidence)})</p>
</div>`;
})
.join('\n');
const listSection = listItems.length > 0 ? ` <ul>\n${listItems.join('\n')}\n </ul>` : '';
const parts = [
`<section class="related-articles" aria-label="${escapeAttr(strings.sectionLabel)}">`,
];
parts.push(` <h3>${escapeText(strings.heading)}</h3>`);
if (listSection) parts.push(listSection);
if (trendBlocks) parts.push(trendBlocks);
parts.push('</section>');
return parts.join('\n');
}
// ─── Private helpers ──────────────────────────────────────────────────────────
/**
* Remove an article ID from every key's list in a lookup map.
* Cleans up empty arrays left behind.
* Skips dangerous keys (`__proto__`, `constructor`, `prototype`) to prevent
* prototype pollution.
* @param map - Lookup map (actor/domain/procedure → article IDs)
* @param keys - Keys to remove the article ID from
* @param articleId - Article ID to remove
*/
function removeIdFromMap(
map: Record<string, string[]>,
keys: readonly string[],
articleId: string
): void {
for (const key of keys) {
Iif (!isSafeKey(key)) continue;
// eslint-disable-next-line security/detect-object-injection -- key validated via isSafeKey
const list = map[key];
Iif (!list) continue;
const filtered = list.filter((id) => id !== articleId);
if (filtered.length === 0) {
// eslint-disable-next-line security/detect-object-injection -- key validated via isSafeKey
delete map[key];
} else E{
// eslint-disable-next-line security/detect-object-injection -- key validated via isSafeKey
map[key] = filtered;
}
}
}
/**
* Add an article ID to every key's list in a lookup map (deduplicating).
* Skips dangerous keys (`__proto__`, `constructor`, `prototype`) to prevent
* prototype pollution.
* @param map - Lookup map (actor/domain/procedure → article IDs)
* @param keys - Keys under which to register the article ID
* @param articleId - Article ID to add
*/
function addIdToMap(
map: Record<string, string[]>,
keys: readonly string[],
articleId: string
): void {
for (const key of keys) {
if (!isSafeKey(key)) continue;
// eslint-disable-next-line security/detect-object-injection -- key validated via isSafeKey
const existing = map[key] ?? [];
if (!existing.includes(articleId)) {
// eslint-disable-next-line security/detect-object-injection -- key validated via isSafeKey
map[key] = [...existing, articleId];
}
}
}
/**
* Convert a string to a URL-safe slug.
*
* Uses Unicode-aware character classes so non-Latin scripts (e.g. Arabic,
* Hebrew, CJK) produce meaningful slugs instead of collapsing to `""`.
* When the result would still be empty (e.g. purely punctuation input) a
* short deterministic hash is returned as a fallback.
*
* @param text - Input string
* @returns Slugified string (never empty)
*/
function slugify(text: string): string {
const slug = text
.toLowerCase()
.replace(/[^\p{L}\p{N}]+/gu, '-')
.replace(/^-+|-+$/g, '');
Eif (slug.length > 0) return slug;
// Deterministic fallback: simple DJB2-style hash of the original text
let hash = 5381;
for (let i = 0; i < text.length; i++) {
hash = ((hash << 5) + hash + text.charCodeAt(i)) | 0;
}
return `h${Math.abs(hash).toString(36)}`;
}
/**
* Escape HTML attribute special characters.
* @param text - Raw text
* @returns Escaped text safe for HTML attributes
*/
function escapeAttr(text: string): string {
return text
.replace(/&/g, '&')
.replace(/"/g, '"')
.replace(/'/g, ''')
.replace(/</g, '<')
.replace(/>/g, '>');
}
/**
* Escape HTML text content special characters.
* @param text - Raw text
* @returns Escaped text safe for HTML text nodes
*/
function escapeText(text: string): string {
return text.replace(/&/g, '&').replace(/</g, '<').replace(/>/g, '>');
}
/** Map from 2-letter language codes to BCP 47 locale tags for date formatting */
const LANG_TO_LOCALE: Record<string, string> = {
en: 'en-GB',
sv: 'sv-SE',
da: 'da-DK',
no: 'nb-NO',
fi: 'fi-FI',
de: 'de-DE',
fr: 'fr-FR',
es: 'es-ES',
nl: 'nl-NL',
ar: 'ar-SA',
he: 'he-IL',
ja: 'ja-JP',
ko: 'ko-KR',
zh: 'zh-CN',
};
/**
* Format an ISO date string as a human-readable date in the given locale.
* @param date - ISO date string (YYYY-MM-DD)
* @param lang - Language code (defaults to 'en')
* @returns Formatted date string
*/
function formatDisplayDate(date: string, lang?: string): string {
const parts = date.split('-');
const year = parts[0] ?? '';
const month = parts[1] ?? '';
const day = parts[2] ?? '';
Iif (!year || !month || !day) return date;
const d = new Date(Date.UTC(parseInt(year, 10), parseInt(month, 10) - 1, parseInt(day, 10)));
const locale = LANG_TO_LOCALE[lang ?? 'en'] ?? 'en-GB';
return d.toLocaleDateString(locale, { day: 'numeric', month: 'short', timeZone: 'UTC' });
}
|