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 | 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 1x 6x 6x 6x 6x 6x 4x 4x 4x 6x 4x 4x 6x 6x 6x 1x 1x 24x 24x 14x 14x 24x 24x 1x 1x 1x 3x 3x 9x 9x 9x 3x 3x 3x 3x 1x 1x 31x 31x 24x 24x 7x 31x 31x 31x 31x 1x 1x 6x 43x 6x 31x 1x 1x 1x 1x 1x 1x 6x 6x 1x 1x 1x 16x 16x 16x 16x 16x 16x 16x 16x 16x 56x 56x 476x 476x 56x 56x 3x 3x 53x 53x 16x 16x 137x 137x 137x 137x 137x 16x 16x 53x 53x 53x 53x 53x 53x 53x 53x 53x 53x 53x 16x 16x 16x 1x 1x 1x 46x 46x 33x 46x 7x 46x 2x 46x 4x 46x 46x 46x 1x 1x 67x 67x 67x 67x 67x 67x 50x 67x 1x 1x 12x 12x 12x 12x 12x 12x 12x 12x 12x 12x 1x 1x 1x 91x 91x 55x 91x 20x 20x 20x 20x 91x 4x 4x 4x 4x 4x 91x 12x 12x 91x 91x 91x 1x 1x 1x 15x 15x 15x 15x 15x 15x 15x 15x 15x 15x 15x 15x 15x 15x 15x 58x 58x 36x 58x 22x 22x 58x 4x 4x 54x 58x 50x 50x 4x 58x 58x 58x 15x 15x 15x 15x 15x 15x 15x 15x 15x 15x 15x 15x 15x 15x 15x 1x 1x 1x 4x 4x 4x 4x 36x 36x 36x 144x 144x 144x 144x 144x 36x 36x 4x 4x 4x 4x 4x 4x 4x 4x 4x 4x 1x 1x 1x 4x 4x 4x 4x 4x 4x 4x 4x 4x 4x 4x 4x 4x 4x 4x 4x 4x 4x 4x 4x 4x 4x 4x 4x 4x 4x 4x 4x 10x 10x 10x 10x 4x 4x 4x 1x 1x 1x 1x 1x 1x | import { FUZZ_SCORER_NAMES, FUZZ_SCORERS } from '@/constants/fuzzScorers';
import type { FuzzScorerName } from '@/constants/fuzzScorers';
import { getAllSupplierQueryCacheEntries, getSearchResults } from '@/utils/idbCache';
import { scoreAstMatch } from '@/utils/search-query/evaluateAst';
import { parseSearchQuery } from '@/utils/search-query/parseSearchQuery';
import type { SearchAst } from '@/utils/search-query/types';
/**
* Dev-console probes for exercising the fuzzy-filter layers against the local
* product cache, with no live search and no network. {@link fuzzTest} compares
* every scorer in `FUZZ_SCORERS` on the plain-query path; {@link astTest} runs
* the advanced boolean path through `parseSearchQuery` → `scoreAstMatch`.
*
* The IO and printing wrappers are kept separate from the pure cores
* ({@link scoreCorpus}, {@link evaluateCorpusAst}) so the scoring logic is
* directly unit-testable.
* @category Utils
* @group Search
* @showCategories
* @categoryDescription Offline fuzzy-filter experimentation helpers.
* @source
*/
/**
* Where a corpus title was read from.
* @category Utils
* @group Types
* @source
*/
export type CorpusSource = 'queryCache' | 'searchResults';
/**
* Which cached stores to build the corpus from. The plain-English aliases
* (`"both"`, `"cache"`, `"results"`) are accepted alongside the store names so the
* console call reads naturally.
* @category Utils
* @group Types
* @source
*/
export type CorpusSourceOption = 'all' | 'both' | 'cache' | 'results' | CorpusSource;
/**
* One cached product title plus the provenance needed to interpret a score.
* @category Utils
* @group Types
* @source
*/
export interface CorpusEntry {
title: string;
supplier?: string;
source: CorpusSource;
/** The query that originally produced this cache entry (query cache only). */
cachedQuery?: string;
url?: string;
}
/**
* Options for {@link fuzzTest} / {@link scoreCorpus}.
* @category Utils
* @group Types
* @source
*/
export interface FuzzProbeOptions {
/** Which caches to draw titles from: `"cache"`, `"results"`, or `"both"`. Default `"both"`. */
source?: CorpusSourceOption;
/** Rows to show/return per scorer. `0` or `Infinity` means all. Default 15. */
limit?: number;
/** Drop titles whose best score across the active scorers is under this. Default 0. */
minScore?: number;
/** One or more suppliers to narrow to; case-insensitive, partial names allowed. */
suppliers?: string | string[];
/** Subset of scorers to run. Default: all of {@link FUZZ_SCORER_NAMES}. */
scorers?: FuzzScorerName[];
/** Print console tables. Default true. */
print?: boolean;
}
/**
* A corpus entry annotated with a single score.
* @category Utils
* @group Types
* @source
*/
export type ScoredEntry = CorpusEntry & { score: number };
/**
* Result of scoring a corpus with every active scorer.
* @category Utils
* @group Types
* @source
*/
export interface FuzzProbeResult {
query: string;
corpusSize: number;
scorers: FuzzScorerName[];
/** Top-`limit` entries per scorer, ranked descending. */
byScorer: Record<string, ScoredEntry[]>;
/** One row per surviving title, with every scorer's score — scorer disagreement at a glance. */
byTitle: Array<CorpusEntry & { scores: Record<string, number> }>;
/** Titles where the scorers disagree most (max − min), descending. */
spread: Array<{ title: string; min: number; max: number; spread: number; avg: number }>;
}
/**
* Options for {@link astTest} / {@link evaluateCorpusAst}.
* @category Utils
* @group Types
* @source
*/
export interface AstProbeOptions {
/** Which caches to draw titles from: `"cache"`, `"results"`, or `"both"`. Default `"both"`. */
source?: CorpusSourceOption;
/** Rows to show per table. `0` or `Infinity` means all. Default 15. */
limit?: number;
/** One or more suppliers to narrow to; case-insensitive, partial names allowed. */
suppliers?: string | string[];
/** Scorer used to rank survivors. Does not affect which titles survive. Default `token_set_ratio`. */
scorer?: FuzzScorerName;
/** Relevance floor passed to `scoreAstMatch`. Default 50 (SupplierBase's `minMatchPercentage`). */
threshold?: number;
/** Per-word fuzzy presence in the leaf gate. Default true (matches fuzzing-enabled production). */
fuzzyWords?: boolean;
/** Also score with `fuzzyWords` inverted and diff the survivor sets. Default true. */
compareFuzzyWords?: boolean;
/** Print console tables. Default true. */
print?: boolean;
}
/**
* A corpus entry that failed the AST predicate, with the leaves responsible.
* @category Utils
* @group Types
* @source
*/
export type DroppedEntry = CorpusEntry & { failedTerms: string[] };
/**
* Result of evaluating a corpus against a parsed search query.
* @category Utils
* @group Types
* @source
*/
export interface AstProbeResult {
query: string;
isAdvanced: boolean;
ast: SearchAst;
/** One-line rendering of {@link ast}, e.g. `(sodium AND NOT borohydride)`. */
astText: string;
corpusSize: number;
scorer: FuzzScorerName;
threshold: number;
fuzzyWords: boolean;
/** Titles satisfying the predicate, ranked by score descending. */
matched: ScoredEntry[];
/** Titles the predicate rejected, each with the phrase(s) that weren't present. */
dropped: DroppedEntry[];
/** Titles whose survival flips with `fuzzyWords`. Empty when `compareFuzzyWords` is false. */
fuzzyWordsDelta: Array<CorpusEntry & { onlyWith: 'fuzzyWords' | 'exact' }>;
}
/** Ranking scorer used by the AST probe when the caller doesn't pick one. */
const DEFAULT_AST_SCORER: FuzzScorerName = 'token_set_ratio';
/** Mirrors `SupplierBase.minMatchPercentage`, the production leaf-score floor. */
const DEFAULT_THRESHOLD = 50;
/** Default rows per printed table. */
const DEFAULT_LIMIT = 15;
/**
* Reads a string property off an untyped cache record without asserting its shape.
* @param item - The candidate record, straight out of IndexedDB.
* @param key - The property to read.
* @returns The value when present and a string, otherwise `undefined`.
*/
function readStringField(item: unknown, key: string): string | undefined {
if (typeof item !== 'object' || item === null || !(key in item)) {
return undefined;
}
const value: unknown = Reflect.get(item, key);
return typeof value === 'string' && value !== '' ? value : undefined;
}
/**
* Builds a deduplicated corpus of product titles from the local IndexedDB caches. No
* network requests are made — this reads `supplier_query_cache` (the raw per-supplier
* result sets, which carry the originating query) and the persisted `search_results`
* row (the products currently in the results table).
* @category Utils
* @param source - Which stores to read: `"cache"` (supplier query cache), `"results"`
* (the current search results), or `"both"`. Defaults to `"both"`.
* @returns The unique cached titles, each tagged with its supplier and origin store.
* @example
* ```typescript
* const corpus = await collectCachedTitles();
* corpus.length; // => 1842
* corpus[0]; // => { title: "Sodium Chloride ACS", supplier: "Loudwolf", source: "queryCache", cachedQuery: "sodium chloride" }
* ```
* @source
*/
export async function collectCachedTitles(
source: CorpusSourceOption = 'all',
): Promise<CorpusEntry[]> {
const entries: CorpusEntry[] = [];
const wantsCache = source === 'all' || source === 'both' || source === 'cache';
const wantsResults = source === 'all' || source === 'both' || source === 'results';
if (wantsCache || source === 'queryCache') {
const cached = await getAllSupplierQueryCacheEntries();
for (const entry of cached) {
// The store is typed `unknown[]` but holds whatever was written, so a legacy
// or partially-written row can be a non-array. Skip it rather than throwing.
if (!Array.isArray(entry.data)) {
continue;
}
for (const item of entry.data) {
const title = readStringField(item, 'title');
if (title === undefined) {
continue;
}
entries.push({
title,
supplier: readStringField(item, 'supplier') ?? entry.__cacheMetadata.supplier,
source: 'queryCache',
cachedQuery: entry.__cacheMetadata.query,
url: readStringField(item, 'url'),
});
}
}
}
if (wantsResults || source === 'searchResults') {
for (const product of await getSearchResults()) {
if (product.title === '') {
continue;
}
entries.push({
title: product.title,
supplier: product.supplier,
source: 'searchResults',
url: product.url,
});
}
}
const seen = new Set<string>();
return entries.filter((entry) => {
const key = `${entry.supplier ?? ''}::${entry.title.toLowerCase()}`;
if (seen.has(key)) {
return false;
}
seen.add(key);
return true;
});
}
/**
* Normalizes the `limit` option, treating `0` and a missing/negative value's
* counterpart sensibly: `0` and `Infinity` both mean "show everything".
* @param limit - The caller's requested row cap.
* @returns A concrete slice length.
*/
function resolveLimit(limit?: number): number {
if (limit === undefined) {
return DEFAULT_LIMIT;
}
return limit <= 0 || !Number.isFinite(limit) ? Number.MAX_SAFE_INTEGER : Math.trunc(limit);
}
/**
* Lists the distinct supplier names present in a corpus, sorted with the most
* titles first. Use it to discover the exact spellings the `suppliers` filter expects.
* @param corpus - The entries to summarize, from {@link collectCachedTitles}.
* @returns Each supplier with its title count, descending.
* @example
* ```typescript
* listSuppliers(await collectCachedTitles());
* // => [{ supplier: "Loudwolf", titles: 412 }, { supplier: "Onyxmet", titles: 118 }]
* ```
* @category Utils
* @source
*/
export function listSuppliers(corpus: CorpusEntry[]): Array<{ supplier: string; titles: number }> {
const counts = new Map<string, number>();
for (const entry of corpus) {
const name = entry.supplier ?? '(unknown)';
counts.set(name, (counts.get(name) ?? 0) + 1);
}
return [...counts]
.map(([supplier, titles]) => ({ supplier, titles }))
.sort((a, b) => b.titles - a.titles);
}
/**
* Narrows a corpus to one or more suppliers. Matching is case-insensitive and
* accepts a partial name, so `"loud"` finds `"Loudwolf"` without the caller having
* to know the exact spelling. Warns (rather than silently returning nothing) when a
* requested supplier matches no cached titles.
* @param corpus - The entries to filter.
* @param suppliers - One name or a list of them; when omitted the corpus passes through.
* @returns The matching subset.
*/
function filterBySuppliers(corpus: CorpusEntry[], suppliers?: string | string[]): CorpusEntry[] {
if (suppliers === undefined) {
return corpus;
}
// Keep the caller's original spelling for the warning, match on the lowered form.
const needles = (Array.isArray(suppliers) ? suppliers : [suppliers])
.map((name) => ({ requested: name.trim(), lowered: name.trim().toLowerCase() }))
.filter((needle) => needle.lowered !== '');
if (needles.length === 0) {
return corpus;
}
const matchesNeedle = (entry: CorpusEntry, lowered: string): boolean =>
entry.supplier?.toLowerCase().includes(lowered) ?? false;
const unmatched = needles.filter((n) => !corpus.some((entry) => matchesNeedle(entry, n.lowered)));
if (unmatched.length > 0) {
console.warn(
`[fuzzScorerLab] no cached titles for supplier(s): ${unmatched.map((n) => n.requested).join(', ')}.` +
` Available: ${listSuppliers(corpus)
.map((s) => `${s.supplier} (${s.titles})`)
.join(', ')}`,
);
}
return corpus.filter((entry) => needles.some((n) => matchesNeedle(entry, n.lowered)));
}
/**
* Scores every corpus title with every active fuzz scorer, so the scorers can be
* compared side by side against real product titles. Pure — performs no IO and prints
* nothing; {@link fuzzTest} is the console-facing wrapper.
* @category Utils
* @param query - The search string to score titles against.
* @param corpus - The titles to score, from {@link collectCachedTitles}.
* @param options - Scorer subset, supplier filter, `minScore` cutoff, and row `limit`.
* @returns Per-scorer rankings, a merged title-by-scorer matrix, and the titles the
* scorers disagree on most.
* @example
* ```typescript
* const result = scoreCorpus("sodium chloride", corpus, { limit: 2 });
* result.byScorer.ratio; // => [{ title: "Sodium Chloride", score: 100, ... }, ...]
* result.spread[0]; // => { title: "NaCl 500g", min: 8, max: 90, spread: 82, avg: 41 }
* ```
* @source
*/
export function scoreCorpus(
query: string,
corpus: CorpusEntry[],
options: FuzzProbeOptions = {},
): FuzzProbeResult {
const scorers = options.scorers ?? [...FUZZ_SCORER_NAMES];
const limit = resolveLimit(options.limit);
const minScore = options.minScore ?? 0;
const filtered = filterBySuppliers(corpus, options.suppliers);
const byTitle: FuzzProbeResult['byTitle'] = [];
for (const entry of filtered) {
const scores: Record<string, number> = {};
for (const name of scorers) {
scores[name] = FUZZ_SCORERS[name](query, entry.title);
}
const values = Object.values(scores);
if (values.length > 0 && Math.max(...values) < minScore) {
continue;
}
byTitle.push({ ...entry, scores });
}
const byScorer: Record<string, ScoredEntry[]> = {};
for (const name of scorers) {
byScorer[name] = byTitle
.map(({ scores, ...entry }) => ({ ...entry, score: scores[name] }))
.sort((a, b) => b.score - a.score)
.slice(0, limit);
}
const spread = byTitle
.map((row) => {
const values = Object.values(row.scores);
const min = Math.min(...values);
const max = Math.max(...values);
return {
title: row.title,
min,
max,
spread: max - min,
avg: Math.round(values.reduce((sum, value) => sum + value, 0) / values.length),
};
})
.sort((a, b) => b.spread - a.spread);
return { query, corpusSize: filtered.length, scorers, byScorer, byTitle, spread };
}
/**
* Renders a parsed search AST as a readable one-line expression, so a printed probe
* shows how the query was actually parsed rather than just echoing the raw input.
* @category Utils
* @param ast - The node to render.
* @returns A parenthesized boolean expression.
* @example
* ```typescript
* renderAst({ type: "and",
* left: { type: "term", value: "sodium", phrase: false },
* right: { type: "not", operand: { type: "term", value: "borohydride", phrase: false } } });
* // => '(sodium AND NOT borohydride)'
* ```
* @source
*/
export function renderAst(ast: SearchAst): string {
switch (ast.type) {
case 'term':
return ast.phrase ? `"${ast.value}"` : ast.value;
case 'and':
return `(${renderAst(ast.left)} AND ${renderAst(ast.right)})`;
case 'or':
return `(${renderAst(ast.left)} OR ${renderAst(ast.right)})`;
case 'not':
return `NOT ${renderAst(ast.operand)}`;
default:
return '<unknown>';
}
}
/**
* Tests whether every word of a phrase is present in a title, replicating the leaf
* gate in `evaluateLeaf` (substring match, or a high-confidence `partial_ratio` hit
* when `fuzzyWords` is on). Used only to explain drops — scoring stays in `scoreAstMatch`.
* @param title - The product title to test.
* @param value - The phrase to look for.
* @param fuzzyWords - Whether per-word fuzzy presence counts.
* @returns True when every word of `value` is present in `title`.
*/
function isPhrasePresent(title: string, value: string, fuzzyWords: boolean): boolean {
const lowerTitle = title.toLowerCase();
const words = value.toLowerCase().split(/\s+/).filter(Boolean);
return words.every(
(word) =>
lowerTitle.includes(word) ||
(fuzzyWords && FUZZ_SCORERS.partial_ratio(word, lowerTitle) >= 88),
);
}
/**
* Whether a title satisfies an AST's predicate, using the same presence gate as
* {@link isPhrasePresent}. Mirrors `scoreAstMatch`'s null/non-null decision without
* computing scores, so {@link explainDrop} can tell a satisfied `NOT` from a missing term.
* @param title - The title to test.
* @param ast - The subtree to evaluate.
* @param fuzzyWords - Whether per-word fuzzy presence counts.
* @returns True when the predicate holds.
*/
function matchesAst(title: string, ast: SearchAst, fuzzyWords: boolean): boolean {
switch (ast.type) {
case 'term':
return isPhrasePresent(title, ast.value, fuzzyWords);
case 'and':
return matchesAst(title, ast.left, fuzzyWords) && matchesAst(title, ast.right, fuzzyWords);
case 'or':
return matchesAst(title, ast.left, fuzzyWords) || matchesAst(title, ast.right, fuzzyWords);
case 'not':
return !matchesAst(title, ast.operand, fuzzyWords);
default:
return false;
}
}
/**
* Explains why a title failed an AST predicate: missing phrases are listed verbatim,
* and a `NOT` branch that excluded the title is reported as `NOT <phrase>`. Every drop
* gets at least one reason, so the dropped table is never blank.
* @category Utils
* @param title - The rejected title.
* @param ast - The parsed query tree.
* @param fuzzyWords - Whether per-word fuzzy presence counts.
* @returns The phrases responsible for the drop.
* @example
* ```typescript
* const { ast } = parseSearchQuery("sodium AND NOT borohydride");
* explainDrop("Potassium Chloride", ast, true); // => ["sodium"]
* explainDrop("Sodium Borohydride 98%", ast, true); // => ["NOT borohydride"]
* ```
* @source
*/
export function explainDrop(title: string, ast: SearchAst, fuzzyWords: boolean): string[] {
switch (ast.type) {
case 'term':
return isPhrasePresent(title, ast.value, fuzzyWords) ? [] : [ast.value];
case 'and':
return [
...explainDrop(title, ast.left, fuzzyWords),
...explainDrop(title, ast.right, fuzzyWords),
];
case 'or': {
const left = explainDrop(title, ast.left, fuzzyWords);
const right = explainDrop(title, ast.right, fuzzyWords);
// An OR only fails when both branches fail; if either matched, nothing is missing.
return left.length > 0 && right.length > 0 ? [...left, ...right] : [];
}
case 'not':
// The NOT held (operand absent), so it isn't the reason; otherwise it is.
return matchesAst(title, ast.operand, fuzzyWords) ? [`NOT ${renderAst(ast.operand)}`] : [];
default:
return [];
}
}
/**
* Evaluates a corpus against an advanced (boolean) search query using the same
* `parseSearchQuery` → `scoreAstMatch` path production suppliers take in
* `SupplierBase.fuzzyFilterAst`. Pure — performs no IO and prints nothing.
*
* Note the leaf gate is word *presence*, not similarity, so the chosen `scorer`
* only ranks survivors; `fuzzyWords` and `threshold` are what change which titles
* survive. `compareFuzzyWords` surfaces exactly that difference.
* @category Utils
* @param query - The raw search string, which may use `AND`/`OR`/`NOT`/parentheses.
* @param corpus - The titles to evaluate, from {@link collectCachedTitles}.
* @param options - Scorer, threshold, `fuzzyWords`, supplier filter, and row `limit`.
* @returns The parsed AST, the ranked survivors, the drops with their failed terms,
* and the titles whose survival depends on `fuzzyWords`.
* @example
* ```typescript
* const result = evaluateCorpusAst("sodium AND NOT borohydride", corpus);
* result.astText; // => '(sodium AND NOT borohydride)'
* result.matched[0]; // => { title: "Sodium Chloride", score: 86, ... }
* result.dropped[0].failedTerms; // => ["sodium"]
* ```
* @source
*/
export function evaluateCorpusAst(
query: string,
corpus: CorpusEntry[],
options: AstProbeOptions = {},
): AstProbeResult {
const scorerName = options.scorer ?? DEFAULT_AST_SCORER;
const threshold = options.threshold ?? DEFAULT_THRESHOLD;
const fuzzyWords = options.fuzzyWords ?? true;
const filtered = filterBySuppliers(corpus, options.suppliers);
const parsed = parseSearchQuery(query);
const evalOptions = { scorer: FUZZ_SCORERS[scorerName], threshold, fuzzyWords };
const compare = options.compareFuzzyWords !== false;
const inverted = { ...evalOptions, fuzzyWords: !fuzzyWords };
const matched: ScoredEntry[] = [];
const dropped: DroppedEntry[] = [];
const fuzzyWordsDelta: AstProbeResult['fuzzyWordsDelta'] = [];
for (const entry of filtered) {
const score = scoreAstMatch(entry.title, parsed.ast, evalOptions);
if (score === null) {
dropped.push({ ...entry, failedTerms: explainDrop(entry.title, parsed.ast, fuzzyWords) });
} else {
matched.push({ ...entry, score });
}
if (!compare) {
continue;
}
const survivesInverted = scoreAstMatch(entry.title, parsed.ast, inverted) !== null;
if ((score !== null) === survivesInverted) {
continue;
}
// Name the fuzzyWords setting that kept it, whichever pass that was.
const keptByFuzzy = score !== null ? fuzzyWords : !fuzzyWords;
fuzzyWordsDelta.push({ ...entry, onlyWith: keptByFuzzy ? 'fuzzyWords' : 'exact' });
}
matched.sort((a, b) => b.score - a.score);
return {
query,
isAdvanced: parsed.isAdvanced,
ast: parsed.ast,
astText: renderAst(parsed.ast),
corpusSize: filtered.length,
scorer: scorerName,
threshold,
fuzzyWords,
matched,
dropped,
fuzzyWordsDelta,
};
}
/**
* Prints a {@link FuzzProbeResult} as console tables: one ranked table per scorer,
* then the merged title-by-scorer matrix, then the biggest scorer disagreements.
* @category Utils
* @param result - The scored result to print.
* @param limit - Rows to show in the merged and spread tables.
* @returns Nothing; writes to the console.
* @example
* ```typescript
* printFuzzProbe(scoreCorpus("acetone", corpus), 10);
* ```
* @source
*/
export function printFuzzProbe(result: FuzzProbeResult, limit?: number): void {
const rows = resolveLimit(limit);
console.info(
`[fuzzTest] query="${result.query}" — ${result.corpusSize} cached titles, ${result.scorers.length} scorers`,
);
for (const name of result.scorers) {
console.groupCollapsed(`${name} — top ${result.byScorer[name].length}`);
console.table(
result.byScorer[name].map((row) => ({
score: row.score,
title: row.title,
supplier: row.supplier,
source: row.source,
})),
);
console.groupEnd();
}
console.groupCollapsed(`All titles × scorers (top ${rows} by average)`);
console.table(
result.byTitle
.map((row) => ({ title: row.title, supplier: row.supplier, ...row.scores }))
.slice(0, rows),
);
console.groupEnd();
console.groupCollapsed(`Biggest scorer disagreements (top ${rows})`);
console.table(result.spread.slice(0, rows));
console.groupEnd();
}
/**
* Prints an {@link AstProbeResult} as console tables: the parsed AST, the ranked
* survivors, the drops with their failed terms, and any `fuzzyWords` flips.
* @category Utils
* @param result - The evaluated result to print.
* @param limit - Rows to show per table.
* @returns Nothing; writes to the console.
* @example
* ```typescript
* printAstProbe(evaluateCorpusAst("acid OR base", corpus), 10);
* ```
* @source
*/
export function printAstProbe(result: AstProbeResult, limit?: number): void {
const rows = resolveLimit(limit);
console.info(
[
`[astTest] query="${result.query}"`,
` parsed: ${result.astText}${result.isAdvanced ? '' : ' (plain query — single term, not advanced syntax)'}`,
` options: scorer=${result.scorer} threshold=${result.threshold} fuzzyWords=${result.fuzzyWords}`,
` corpus: ${result.corpusSize} titles → ${result.matched.length} matched, ${result.dropped.length} dropped`,
` note: the scorer only ranks survivors; fuzzyWords/threshold decide who survives`,
].join('\n'),
);
// A NOT branch scores exactly `threshold`, and AND takes the min — so an
// `X AND NOT Y` query pins every survivor to the floor and the ranking is
// meaningless. Worth saying out loud rather than letting it look like a tie.
const distinctScores = new Set(result.matched.map((row) => row.score));
if (result.matched.length > 1 && distinctScores.size === 1) {
console.warn(
`[astTest] every survivor scored ${[...distinctScores][0]} — ranking is degenerate.` +
` A NOT branch contributes exactly the threshold and AND takes the min, so` +
` "X AND NOT Y" flattens all scores to the floor.`,
);
}
console.groupCollapsed(
`Matched — top ${Math.min(rows, result.matched.length)} of ${result.matched.length}`,
);
console.table(
result.matched.slice(0, rows).map((row) => ({
score: row.score,
title: row.title,
supplier: row.supplier,
source: row.source,
})),
);
console.groupEnd();
console.groupCollapsed(
`Dropped — top ${Math.min(rows, result.dropped.length)} of ${result.dropped.length}`,
);
console.table(
result.dropped.slice(0, rows).map((row) => ({
title: row.title,
supplier: row.supplier,
failedTerms: row.failedTerms.join(', '),
})),
);
console.groupEnd();
if (result.fuzzyWordsDelta.length > 0) {
console.groupCollapsed(`fuzzyWords flips — ${result.fuzzyWordsDelta.length} titles`);
console.table(
result.fuzzyWordsDelta.slice(0, rows).map((row) => ({
title: row.title,
supplier: row.supplier,
onlyWith: row.onlyWith,
})),
);
console.groupEnd();
}
}
/**
* Runs a query against every cached product title with all nine fuzz scorers and
* prints them side by side — the offline equivalent of `SupplierBase.fuzzyFilter`'s
* dev-only comparison table, but over the whole local cache and with no search.
* @category Utils
* @param query - The search string to score cached titles against.
* @param options - Corpus source, supplier filter, scorer subset, `minScore`, `rows`,
* and whether to print.
* @returns The full scored result, also printed unless `print: false`.
* @example
* ```typescript
* // In the console:
* const result = await chempal.fuzzTest("sodium borohydride");
* // ┌─────────┬───────┬──────────────────────────────┬────────────┐
* // │ (index) │ score │ title │ supplier │
* // ├─────────┼───────┼──────────────────────────────┼────────────┤
* // │ 0 │ 100 │ 'Sodium Borohydride 98%' │ 'Loudwolf' │
* // └─────────┴───────┴──────────────────────────────┴────────────┘
* result.spread[0]; // the title the scorers disagree on most
*
* await chempal.fuzzTest("acetone", { supplier: "Loudwolf", rows: 25 });
* ```
* @source
*/
export async function fuzzTest(
query: string,
options: FuzzProbeOptions = {},
): Promise<FuzzProbeResult> {
const corpus = await collectCachedTitles(options.source);
const result = scoreCorpus(query, corpus, options);
if (options.print !== false) {
printFuzzProbe(result, options.limit);
}
return result;
}
/**
* Runs an advanced (boolean) query against every cached product title through the
* real `parseSearchQuery` → `scoreAstMatch` path, showing which titles the predicate
* keeps, which it drops and why, and which drops hinge on `fuzzyWords`.
* @category Utils
* @param query - The search string; may use `AND`/`OR`/`NOT` and parentheses.
* @param options - `source` (`"cache"` / `"results"` / `"both"`), `suppliers`, `limit`
* (`0` = all), ranking `scorer`, `threshold`, `fuzzyWords`, and `compareFuzzyWords`.
* @returns The full evaluated result, also printed unless `print: false`.
* @example
* ```typescript
* // In the console:
* const result = await chempal.astTest("sodium AND NOT borohydride");
* result.astText; // => '(sodium AND NOT borohydride)'
* result.matched.length; // => 37
* result.dropped[0].failedTerms; // => ['sodium']
*
* await chempal.astTest("acid OR base", { fuzzyWords: false, threshold: 70 });
* await chempal.astTest("sodium", { suppliers: "loud", source: "cache", limit: 0 });
* ```
* @source
*/
export async function astTest(
query: string,
options: AstProbeOptions = {},
): Promise<AstProbeResult> {
const corpus = await collectCachedTitles(options.source);
const result = evaluateCorpusAst(query, corpus, options);
if (options.print !== false) {
printAstProbe(result, options.limit);
}
return result;
}
|