@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
Support "-term" in Ferret engine queries
Summary:
Ref T12819. Supports negating search terms, e.g. "apple -honeycrisp".
When negating a term, we're a little more strict about what can match (that is, what can //prevent// a document from being returned) since it's easy for a user to type "apple -honeycrisp -honey -crisp -crispies -olcrispers -honeyyums" to keep refining their search, but hard/impossible to split apart an overboard term.
Test Plan:
- Ran `apple -smith`, `apple -"granny smith"`, etc.
- Verified `phone -tact` does not exclude `phone contact`.
- (In theory, `phone -~tact` would, but the parser currently doesn't support this, and I'm not champing at the bit to add support.)
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T12819
Differential Revision: https://secure.phabricator.com/D18502
+48
-9
+48
-9
src/infrastructure/query/policy/PhabricatorCursorPagedPolicyAwareQuery.php
+48
-9
src/infrastructure/query/policy/PhabricatorCursorPagedPolicyAwareQuery.php
···
1410
1410
}
1411
1411
1412
1412
$op_sub = PhutilSearchQueryCompiler::OPERATOR_SUBSTRING;
1413
+
$op_not = PhutilSearchQueryCompiler::OPERATOR_NOT;
1413
1414
1414
1415
$engine = $this->ferretEngine;
1415
1416
$ngram_engine = new PhabricatorNgramEngine();
···
1421
1422
$flat = array();
1422
1423
foreach ($this->ferretTokens as $fulltext_token) {
1423
1424
$raw_token = $fulltext_token->getToken();
1425
+
1426
+
// If this is a negated term like "-pomegranate", don't join the ngram
1427
+
// table since we aren't looking for documents with this term. (We could
1428
+
// LEFT JOIN the table and require a NULL row, but this is probably more
1429
+
// trouble than it's worth.)
1430
+
if ($raw_token->getOperator() == $op_not) {
1431
+
continue;
1432
+
}
1433
+
1424
1434
$value = $raw_token->getValue();
1425
1435
1426
1436
$length = count(phutil_utf8v($value));
···
1530
1540
1531
1541
$ngram_engine = new PhabricatorNgramEngine();
1532
1542
$stemmer = new PhutilSearchStemmer();
1543
+
1533
1544
$op_sub = PhutilSearchQueryCompiler::OPERATOR_SUBSTRING;
1545
+
$op_not = PhutilSearchQueryCompiler::OPERATOR_NOT;
1534
1546
1535
1547
$where = array();
1536
1548
foreach ($this->ferretTokens as $fulltext_token) {
1537
1549
$raw_token = $fulltext_token->getToken();
1538
1550
$value = $raw_token->getValue();
1551
+
1552
+
$is_not = ($raw_token->getOperator() == $op_not);
1539
1553
1540
1554
if ($raw_token->getOperator() == $op_sub) {
1541
1555
$is_substring = true;
···
1546
1560
// If we're doing substring search, we just match against the raw corpus
1547
1561
// and we're done.
1548
1562
if ($is_substring) {
1549
-
$where[] = qsprintf(
1550
-
$conn,
1551
-
'(ftfield.rawCorpus LIKE %~)',
1552
-
$value);
1563
+
if ($is_not) {
1564
+
$where[] = qsprintf(
1565
+
$conn,
1566
+
'(ftfield.rawCorpus NOT LIKE %~)',
1567
+
$value);
1568
+
} else {
1569
+
$where[] = qsprintf(
1570
+
$conn,
1571
+
'(ftfield.rawCorpus LIKE %~)',
1572
+
$value);
1573
+
}
1553
1574
continue;
1554
1575
}
1555
1576
···
1563
1584
$is_stemmed = true;
1564
1585
}
1565
1586
1587
+
// Never stem negated queries, since this can exclude results users
1588
+
// did not mean to exclude and generally confuse things.
1589
+
if ($is_not) {
1590
+
$is_stemmed = false;
1591
+
}
1592
+
1566
1593
$term_constraints = array();
1567
1594
1568
1595
$term_value = ' '.$ngram_engine->newTermsCorpus($value).' ';
1569
-
$term_constraints[] = qsprintf(
1570
-
$conn,
1571
-
'(ftfield.termCorpus LIKE %~)',
1572
-
$term_value);
1596
+
if ($is_not) {
1597
+
$term_constraints[] = qsprintf(
1598
+
$conn,
1599
+
'(ftfield.termCorpus NOT LIKE %~)',
1600
+
$term_value);
1601
+
} else {
1602
+
$term_constraints[] = qsprintf(
1603
+
$conn,
1604
+
'(ftfield.termCorpus LIKE %~)',
1605
+
$term_value);
1606
+
}
1573
1607
1574
1608
if ($is_stemmed) {
1575
1609
$stem_value = $stemmer->stemToken($value);
···
1582
1616
$stem_value);
1583
1617
}
1584
1618
1585
-
if ($is_quoted) {
1619
+
if ($is_not) {
1620
+
$where[] = qsprintf(
1621
+
$conn,
1622
+
'(%Q)',
1623
+
implode(' AND ', $term_constraints));
1624
+
} else if ($is_quoted) {
1586
1625
$where[] = qsprintf(
1587
1626
$conn,
1588
1627
'(ftfield.rawCorpus LIKE %~ AND (%Q))',