@recaptime-dev's working patches + fork for Phorge, a community fork of Phabricator. (Upstream dev and stable branches are at upstream/main and upstream/stable respectively.)
hq.recaptime.dev/wiki/Phorge
phorge
phabricator
Add support for relevance-ranking Ferret engine results
Summary: Ref T12819. "Relevance" here just means "how many of your search terms are present in the title?" but that's about the best we can do anyway.
Test Plan: Indexed tasks "A B", "A Z", "Z B", and "Z Z" (all with "A B" in comments). Searched for "A B". Got results ranked in the listed order, with "A B" as the most relevant hit for query "A B".
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T12819
Differential Revision: https://secure.phabricator.com/D18539
+120
-1
+120
-1
src/infrastructure/query/policy/PhabricatorCursorPagedPolicyAwareQuery.php
+120
-1
src/infrastructure/query/policy/PhabricatorCursorPagedPolicyAwareQuery.php
···
251
251
}
252
252
253
253
$select[] = $this->buildEdgeLogicSelectClause($conn);
254
+
$select[] = $this->buildFerretSelectClause($conn);
254
255
255
256
return $select;
256
257
}
···
769
770
}
770
771
}
771
772
773
+
if ($this->supportsFerretEngine()) {
774
+
$orders['relevance'] = array(
775
+
'vector' => array('rank', 'id'),
776
+
'name' => pht('Relevence'),
777
+
);
778
+
}
779
+
772
780
return $orders;
773
781
}
774
782
···
959
967
'customfield.index.key' => $digest,
960
968
);
961
969
}
970
+
}
971
+
972
+
if ($this->supportsFerretEngine()) {
973
+
$columns['rank'] = array(
974
+
'table' => null,
975
+
'column' => '_ft_rank',
976
+
'type' => 'int',
977
+
);
962
978
}
963
979
964
980
$cache->setKey($cache_key, $columns);
···
1385
1401
/* -( Ferret )------------------------------------------------------------- */
1386
1402
1387
1403
1404
+
public function supportsFerretEngine() {
1405
+
$object = $this->newResultObject();
1406
+
return ($object instanceof PhabricatorFerretInterface);
1407
+
}
1408
+
1409
+
1388
1410
public function withFerretConstraint(
1389
1411
PhabricatorFerretEngine $engine,
1390
1412
array $fulltext_tokens) {
1413
+
1414
+
if (!$this->supportsFerretEngine()) {
1415
+
throw new Exception(
1416
+
pht(
1417
+
'Query ("%s") does not support the Ferret fulltext engine.',
1418
+
get_class($this)));
1419
+
}
1391
1420
1392
1421
if ($this->ferretEngine) {
1393
1422
throw new Exception(
···
1416
1445
$raw_field = $engine->getFieldForFunction($function);
1417
1446
1418
1447
if (!isset($table_map[$function])) {
1419
-
$alias = 'ftfield'.$idx++;
1448
+
$alias = 'ftfield_'.$idx++;
1420
1449
$table_map[$function] = array(
1421
1450
'alias' => $alias,
1422
1451
'key' => $raw_field,
···
1426
1455
$current_function = $function;
1427
1456
}
1428
1457
1458
+
// Join the title field separately so we can rank results.
1459
+
$table_map['rank'] = array(
1460
+
'alias' => 'ft_rank',
1461
+
'key' => PhabricatorSearchDocumentFieldType::FIELD_TITLE,
1462
+
);
1463
+
1429
1464
$this->ferretTables = $table_map;
1430
1465
1431
1466
return $this;
1467
+
}
1468
+
1469
+
protected function buildFerretSelectClause(AphrontDatabaseConnection $conn) {
1470
+
$select = array();
1471
+
1472
+
if (!$this->supportsFerretEngine()) {
1473
+
return $select;
1474
+
}
1475
+
1476
+
if (!$this->ferretEngine) {
1477
+
$select[] = '0 _ft_rank';
1478
+
return $select;
1479
+
}
1480
+
1481
+
$engine = $this->ferretEngine;
1482
+
$stemmer = $engine->newStemmer();
1483
+
1484
+
$op_sub = PhutilSearchQueryCompiler::OPERATOR_SUBSTRING;
1485
+
$op_not = PhutilSearchQueryCompiler::OPERATOR_NOT;
1486
+
$table_alias = 'ft_rank';
1487
+
1488
+
$parts = array();
1489
+
foreach ($this->ferretTokens as $fulltext_token) {
1490
+
$raw_token = $fulltext_token->getToken();
1491
+
$value = $raw_token->getValue();
1492
+
1493
+
if ($raw_token->getOperator() == $op_not) {
1494
+
// Ignore "not" terms when ranking, since they aren't useful.
1495
+
continue;
1496
+
}
1497
+
1498
+
if ($raw_token->getOperator() == $op_sub) {
1499
+
$is_substring = true;
1500
+
} else {
1501
+
$is_substring = false;
1502
+
}
1503
+
1504
+
if ($is_substring) {
1505
+
$parts[] = qsprintf(
1506
+
$conn,
1507
+
'IF(%T.rawCorpus LIKE %~, 2, 0)',
1508
+
$table_alias,
1509
+
$value);
1510
+
continue;
1511
+
}
1512
+
1513
+
if ($raw_token->isQuoted()) {
1514
+
$is_quoted = true;
1515
+
$is_stemmed = false;
1516
+
} else {
1517
+
$is_quoted = false;
1518
+
$is_stemmed = true;
1519
+
}
1520
+
1521
+
$term_constraints = array();
1522
+
1523
+
$term_value = $engine->newTermsCorpus($value);
1524
+
1525
+
$parts[] = qsprintf(
1526
+
$conn,
1527
+
'IF(%T.termCorpus LIKE %~, 2, 0)',
1528
+
$table_alias,
1529
+
$term_value);
1530
+
1531
+
if ($is_stemmed) {
1532
+
$stem_value = $stemmer->stemToken($value);
1533
+
$stem_value = $engine->newTermsCorpus($stem_value);
1534
+
1535
+
$parts[] = qsprintf(
1536
+
$conn,
1537
+
'IF(%T.normalCorpus LIKE %~, 1, 0)',
1538
+
$table_alias,
1539
+
$stem_value);
1540
+
}
1541
+
1542
+
$parts[] = '0';
1543
+
}
1544
+
1545
+
$select[] = qsprintf(
1546
+
$conn,
1547
+
'%Q _ft_rank',
1548
+
implode(' + ', $parts));
1549
+
1550
+
return $select;
1432
1551
}
1433
1552
1434
1553
protected function buildFerretJoinClause(AphrontDatabaseConnection $conn) {