Add SI-prefix-aware sorting column for parts tableFeature/si value sort (#1344)

* Add SI-prefix-aware sorting column for the parts table

Adds an optional "Name (SI)" column that parses numeric values with SI
prefixes (p, n, u/µ, m, k/K, M, G, T) from part names and sorts by the
resulting physical value. This is useful for electronic components where
alphabetical sorting produces wrong results — e.g. 100nF, 10pF, 1uF
should sort as 10pF < 100nF < 1uF.

Implementation:
- New SiValueSort DQL function with platform-specific SQL generation
  for PostgreSQL (POSIX regex), MySQL/MariaDB (REGEXP_SUBSTR), and
  SQLite (PHP callback registered via the existing middleware).
- The regex is start-anchored: only names beginning with a number are
  matched. Part numbers like "MCP2515" or "Crystal 20MHz" are ignored.
- When SI sort is active, NATSORT is appended as a secondary sort so
  that non-matching parts fall back to natural string ordering instead
  of appearing in arbitrary order.
- The column is opt-in (not in default columns) and displays the parsed
  float value, or an empty cell for non-matching names.

* Rename SI column from "Name (SI)" to "SI Value"

The column now shows the parsed numeric value rather than the part name,
so the label should reflect that.

* Support comma as decimal separator in SI value parsing

Part names using European decimal notation (e.g. "4,7 kΩ", "2,2uF")
were parsed incorrectly because the regex only recognized dots. Now
commas are normalized to dots before parsing, matching the existing
pattern used elsewhere in the codebase (PartNormalizer, price providers).
This commit is contained in:
Wieland Schopohl 2026-04-15 22:56:34 +02:00 committed by GitHub
parent 146e85f84c
commit 29db029d69
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
7 changed files with 427 additions and 0 deletions

View file

@ -56,6 +56,7 @@ doctrine:
natsort: App\Doctrine\Functions\Natsort
array_position: App\Doctrine\Functions\ArrayPosition
ilike: App\Doctrine\Functions\ILike
si_value_sort: App\Doctrine\Functions\SiValueSort
when@test:
doctrine:

View file

@ -38,6 +38,7 @@ use App\DataTables\Filters\PartFilter;
use App\DataTables\Filters\PartSearchFilter;
use App\DataTables\Helpers\ColumnSortHelper;
use App\DataTables\Helpers\PartDataTableHelper;
use App\Doctrine\Functions\SiValueSort;
use App\Doctrine\Helpers\FieldHelper;
use App\Entity\Parts\ManufacturingStatus;
use App\Entity\Parts\Part;
@ -118,6 +119,17 @@ final class PartsDataTable implements DataTableTypeInterface
'render' => fn($value, Part $context) => $this->partDataTableHelper->renderName($context),
'orderField' => 'NATSORT(part.name)'
])
->add('si_value', TextColumn::class, [
'label' => $this->translator->trans('part.table.si_value'),
'render' => function ($value, Part $context): string {
$siValue = SiValueSort::sqliteSiValue($context->getName());
if ($siValue !== null) {
return htmlspecialchars(sprintf('%g', $siValue));
}
return '';
},
'orderField' => 'SI_VALUE_SORT(part.name)',
])
->add('id', TextColumn::class, [
'label' => $this->translator->trans('part.table.id'),
])
@ -484,6 +496,19 @@ final class PartsDataTable implements DataTableTypeInterface
//$builder->addGroupBy('_bulkImportJob');
}
//When sorting by SI value, add NATSORT as a secondary sort so that parts without
//an SI-prefixed value fall back to natural string ordering seamlessly.
$orderByParts = $builder->getDQLPart('orderBy');
foreach ($orderByParts as $orderBy) {
foreach ($orderBy->getParts() as $part) {
if (str_contains($part, 'SI_VALUE_SORT')) {
$direction = str_contains($part, 'DESC') ? 'DESC' : 'ASC';
$builder->addOrderBy('NATSORT(part.name)', $direction);
break 2;
}
}
}
return $builder;
}

View file

@ -0,0 +1,196 @@
<?php
/*
* This file is part of Part-DB (https://github.com/Part-DB/Part-DB-symfony).
*
* Copyright (C) 2019 - 2024 Jan Böhmer (https://github.com/jbtronics)
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU Affero General Public License as published
* by the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU Affero General Public License for more details.
*
* You should have received a copy of the GNU Affero General Public License
* along with this program. If not, see <https://www.gnu.org/licenses/>.
*/
declare(strict_types=1);
namespace App\Doctrine\Functions;
use Doctrine\DBAL\Platforms\AbstractMySQLPlatform;
use Doctrine\DBAL\Platforms\PostgreSQLPlatform;
use Doctrine\DBAL\Platforms\SQLitePlatform;
use Doctrine\ORM\Query\AST\Functions\FunctionNode;
use Doctrine\ORM\Query\AST\Node;
use Doctrine\ORM\Query\Parser;
use Doctrine\ORM\Query\SqlWalker;
use Doctrine\ORM\Query\TokenType;
/**
* Custom DQL function that extracts the first numeric value with an optional SI prefix
* from a string and returns the scaled numeric value for sorting.
*
* Usage: SI_VALUE_SORT(part.name)
*
* This enables sorting parts by their physical value. For example, capacitors
* named "100nF", "1uF", "10pF" will be sorted by actual value: 10pF < 100nF < 1uF.
*
* Supported SI prefixes: p (pico, 1e-12), n (nano, 1e-9), u/µ (micro, 1e-6),
* m (milli, 1e-3), k/K (kilo, 1e3), M (mega, 1e6), G (giga, 1e9), T (tera, 1e12).
*
* Only matches numbers at the very beginning of the string (ignoring leading whitespace).
* Names like "Crystal 20MHz" will NOT match since the number is not at the start.
* Names without a recognizable numeric+prefix pattern return NULL and sort last.
*/
class SiValueSort extends FunctionNode
{
private ?Node $field = null;
/**
* SI prefix multipliers. Used by the SQLite PHP callback.
*/
private const SI_MULTIPLIERS = [
'p' => 1e-12,
'n' => 1e-9,
'u' => 1e-6,
'µ' => 1e-6,
'm' => 1e-3,
'k' => 1e3,
'K' => 1e3,
'M' => 1e6,
'G' => 1e9,
'T' => 1e12,
];
public function parse(Parser $parser): void
{
$parser->match(TokenType::T_IDENTIFIER);
$parser->match(TokenType::T_OPEN_PARENTHESIS);
$this->field = $parser->ArithmeticExpression();
$parser->match(TokenType::T_CLOSE_PARENTHESIS);
}
public function getSql(SqlWalker $sqlWalker): string
{
assert($this->field !== null, 'Field is not set');
$platform = $sqlWalker->getConnection()->getDatabasePlatform();
$rawField = $this->field->dispatch($sqlWalker);
// Normalize comma decimal separator to dot for SQL platforms (European locale support)
$fieldSql = "REPLACE({$rawField}, ',', '.')";
if ($platform instanceof PostgreSQLPlatform) {
return $this->getPostgreSQLSql($fieldSql);
}
if ($platform instanceof AbstractMySQLPlatform) {
return $this->getMySQLSql($fieldSql);
}
// SQLite: comma normalization is handled in the PHP callback
$fieldSql = $rawField;
if ($platform instanceof SQLitePlatform) {
return "SI_VALUE({$fieldSql})";
}
// Fallback: return NULL (no SI sorting available)
return 'NULL';
}
/**
* PostgreSQL implementation using substring() with POSIX regex.
*/
private function getPostgreSQLSql(string $field): string
{
// Extract the numeric part using POSIX regex, anchored at start (with optional leading whitespace)
$numericPart = "CAST(substring({$field} FROM '^\\s*(\\d+\\.?\\d*)\\s*[pnuµmkKMGT]?') AS DOUBLE PRECISION)";
// Extract the SI prefix character
$prefixPart = "substring({$field} FROM '^\\s*\\d+\\.?\\d*\\s*([pnuµmkKMGT])')";
return $this->buildCaseExpression($numericPart, $prefixPart);
}
/**
* MySQL/MariaDB implementation using REGEXP_SUBSTR.
*/
private function getMySQLSql(string $field): string
{
// Extract the numeric part, anchored at start (with optional leading whitespace)
$numericPart = "CAST(REGEXP_SUBSTR({$field}, '^[[:space:]]*[0-9]+\\.?[0-9]*') AS DECIMAL(30,15))";
// Extract the prefix: get the full number+prefix match anchored at start, then take the last char
$fullMatch = "REGEXP_SUBSTR({$field}, '^[[:space:]]*[0-9]+\\.?[0-9]*[[:space:]]*[pnuµmkKMGT]')";
$prefixPart = "RIGHT({$fullMatch}, 1)";
return $this->buildCaseExpression($numericPart, $prefixPart);
}
/**
* Build a CASE expression that maps an SI prefix character to a multiplier
* and multiplies it with the numeric value.
*
* @param string $numericExpr SQL expression that evaluates to the numeric part
* @param string $prefixExpr SQL expression that evaluates to the SI prefix character
* @return string SQL CASE expression
*/
private function buildCaseExpression(string $numericExpr, string $prefixExpr): string
{
return "(CASE" .
" WHEN {$numericExpr} IS NULL THEN NULL" .
" WHEN {$prefixExpr} = 'p' THEN {$numericExpr} * 1e-12" .
" WHEN {$prefixExpr} = 'n' THEN {$numericExpr} * 1e-9" .
" WHEN {$prefixExpr} = 'u' THEN {$numericExpr} * 1e-6" .
" WHEN {$prefixExpr} = 'µ' THEN {$numericExpr} * 1e-6" .
" WHEN {$prefixExpr} = 'm' THEN {$numericExpr} * 1e-3" .
" WHEN {$prefixExpr} = 'k' THEN {$numericExpr} * 1e3" .
" WHEN {$prefixExpr} = 'K' THEN {$numericExpr} * 1e3" .
" WHEN {$prefixExpr} = 'M' THEN {$numericExpr} * 1e6" .
" WHEN {$prefixExpr} = 'G' THEN {$numericExpr} * 1e9" .
" WHEN {$prefixExpr} = 'T' THEN {$numericExpr} * 1e12" .
" ELSE {$numericExpr} * 1" .
" END)";
}
/**
* PHP callback for SQLite's SI_VALUE function.
* Extracts the first numeric value with an optional SI prefix and returns the scaled value.
*
* @param string|null $value The input string
* @return float|null The scaled numeric value, or null if no number found
*/
public static function sqliteSiValue(?string $value): ?float
{
if ($value === null) {
return null;
}
// Normalize comma decimal separator to dot (European locale support)
$value = str_replace(',', '.', $value);
// Match a number at the very start (allowing leading whitespace), optionally followed by an SI prefix
if (!preg_match('/^\s*(\d+\.?\d*)\s*([pnuµmkKMGT])?/u', $value, $matches)) {
return null;
}
$number = (float) $matches[1];
$prefix = $matches[2] ?? '';
if ($prefix === '') {
return $number;
}
$multiplier = self::SI_MULTIPLIERS[$prefix] ?? 1.0;
return $number * $multiplier;
}
}

View file

@ -23,6 +23,7 @@ declare(strict_types=1);
namespace App\Doctrine\Middleware;
use App\Doctrine\Functions\SiValueSort;
use App\Exceptions\InvalidRegexException;
use Doctrine\DBAL\Driver\Connection;
use Doctrine\DBAL\Driver\Middleware\AbstractDriverMiddleware;
@ -51,6 +52,9 @@ class SQLiteRegexExtensionMiddlewareDriver extends AbstractDriverMiddleware
//Create a new collation for natural sorting
$native_connection->sqliteCreateCollation('NATURAL_CMP', strnatcmp(...));
//Create a function for SI prefix value sorting
$native_connection->sqliteCreateFunction('SI_VALUE', SiValueSort::sqliteSiValue(...), 1, \PDO::SQLITE_DETERMINISTIC);
}
}

View file

@ -52,6 +52,8 @@ enum PartTableColumns : string implements TranslatableInterface
case TAGS = "tags";
case ATTACHMENTS = "attachments";
case SI_VALUE = "si_value";
case EDA_REFERENCE = "eda_reference";
case EDA_VALUE = "eda_value";

View file

@ -0,0 +1,193 @@
<?php
/*
* This file is part of Part-DB (https://github.com/Part-DB/Part-DB-symfony).
*
* Copyright (C) 2019 - 2026 Jan Böhmer (https://github.com/jbtronics)
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU Affero General Public License as published
* by the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU Affero General Public License for more details.
*
* You should have received a copy of the GNU Affero General Public License
* along with this program. If not, see <https://www.gnu.org/licenses/>.
*/
declare(strict_types=1);
namespace App\Tests\Doctrine\Functions;
use App\Doctrine\Functions\SiValueSort;
use Doctrine\DBAL\Platforms\MySQLPlatform;
use Doctrine\DBAL\Platforms\PostgreSQLPlatform;
use Doctrine\DBAL\Platforms\SQLitePlatform;
final class SiValueSortTest extends AbstractDoctrineFunctionTestCase
{
public function testPostgreSQLGeneratesCaseExpression(): void
{
$function = new SiValueSort('SI_VALUE_SORT');
$this->setObjectProperty($function, 'field', $this->createNode('part_name'));
$sql = $function->getSql($this->createSqlWalker(new PostgreSQLPlatform()));
$this->assertStringContainsString('CASE', $sql);
$this->assertStringContainsString("REPLACE(part_name, ',', '.')", $sql);
$this->assertStringContainsString('1e-12', $sql);
$this->assertStringContainsString('1e-9', $sql);
$this->assertStringContainsString('1e-6', $sql);
$this->assertStringContainsString('1e-3', $sql);
$this->assertStringContainsString('1e3', $sql);
$this->assertStringContainsString('1e6', $sql);
$this->assertStringContainsString('1e9', $sql);
$this->assertStringContainsString('1e12', $sql);
}
public function testMySQLGeneratesCaseExpression(): void
{
$function = new SiValueSort('SI_VALUE_SORT');
$this->setObjectProperty($function, 'field', $this->createNode('part_name'));
$sql = $function->getSql($this->createSqlWalker(new MySQLPlatform()));
$this->assertStringContainsString('CASE', $sql);
$this->assertStringContainsString("REPLACE(part_name, ',', '.')", $sql);
$this->assertStringContainsString('1e-12', $sql);
$this->assertStringContainsString('1e6', $sql);
}
public function testSQLiteUsesSiValueFunction(): void
{
$function = new SiValueSort('SI_VALUE_SORT');
$this->setObjectProperty($function, 'field', $this->createNode('part_name'));
$sql = $function->getSql($this->createSqlWalker(new SQLitePlatform()));
$this->assertSame('SI_VALUE(part_name)', $sql);
}
/**
* @dataProvider sqliteSiValueProvider
*/
public function testSqliteSiValue(?string $input, ?float $expected): void
{
$result = SiValueSort::sqliteSiValue($input);
if ($expected === null) {
$this->assertNull($result);
} else {
$this->assertEqualsWithDelta($expected, $result, $expected * 1e-9);
}
}
/**
* @return iterable<string, array{?string, ?float}>
*/
public static function sqliteSiValueProvider(): iterable
{
// Basic SI prefix values
yield 'pico' => ['10pF', 10e-12];
yield 'nano' => ['100nF', 100e-9];
yield 'micro_u' => ['1uF', 1e-6];
yield 'micro_µ' => ['1µF', 1e-6];
yield 'milli' => ['4.7mH', 4.7e-3];
yield 'kilo_lower' => ['4.7k', 4.7e3];
yield 'kilo_upper' => ['4.7K', 4.7e3];
yield 'mega' => ['1M', 1e6];
yield 'giga' => ['2.2G', 2.2e9];
yield 'tera' => ['1T', 1e12];
// No prefix (plain number)
yield 'plain_integer' => ['100', 100.0];
yield 'plain_decimal' => ['4.7', 4.7];
// Decimal values with prefix (dot separator)
yield 'decimal_nano' => ['4.7nF', 4.7e-9];
yield 'decimal_micro' => ['0.1uF', 0.1e-6];
yield 'decimal_kilo' => ['2.2k', 2.2e3];
// Comma decimal separator (European locale)
yield 'comma_kilo' => ['4,7k', 4.7e3];
yield 'comma_micro' => ['2,2uF', 2.2e-6];
yield 'comma_kilo_space' => ['1,2 kΩ', 1.2e3];
// Number NOT at the start — should return NULL
yield 'prefixed_name' => ['CAP-100nF', null];
yield 'name_with_number' => ['R 4.7k 1%', null];
yield 'crystal' => ['Crystal 20MHz', null];
// Number at start with trailing text
yield 'number_with_suffix' => ['10nF 25V', 10e-9];
// Space between number and prefix
yield 'space_before_prefix' => ['100 nF', 100e-9];
// Leading whitespace before number
yield 'leading_whitespace' => [' 10uF', 10e-6];
// No number at all
yield 'no_number' => ['Connector', null];
yield 'text_only' => ['LED red', null];
// Null input
yield 'null' => [null, null];
// Empty string
yield 'empty' => ['', null];
}
/**
* Test that the sort order is correct by comparing sqliteSiValue results.
*/
public function testSortOrder(): void
{
$parts = ['1uF', '100nF', '10pF', '10uF', '0.1mF', '1F', '10kF', '1MF'];
$expected = ['10pF', '100nF', '1uF', '10uF', '0.1mF', '1F', '10kF', '1MF'];
// Sort using sqliteSiValue
usort($parts, static function (string $a, string $b): int {
$va = SiValueSort::sqliteSiValue($a);
$vb = SiValueSort::sqliteSiValue($b);
return $va <=> $vb;
});
$this->assertSame($expected, $parts);
}
/**
* Test that NULL values sort last (after all numeric values).
*/
public function testNullSortsLast(): void
{
$parts = ['Connector', '100nF', 'LED red', '10pF'];
usort($parts, static function (string $a, string $b): int {
$va = SiValueSort::sqliteSiValue($a);
$vb = SiValueSort::sqliteSiValue($b);
// NULL sorts last
if ($va === null && $vb === null) {
return 0;
}
if ($va === null) {
return 1;
}
if ($vb === null) {
return -1;
}
return $va <=> $vb;
});
$this->assertSame('10pF', $parts[0]);
$this->assertSame('100nF', $parts[1]);
// Last two should be the non-numeric names
$this->assertContains('Connector', array_slice($parts, 2));
$this->assertContains('LED red', array_slice($parts, 2));
}
}

View file

@ -2780,6 +2780,12 @@ If you have done this incorrectly or if a computer is no longer trusted, you can
<target>Name</target>
</segment>
</unit>
<unit id="sIvAlUe" name="part.table.si_value">
<segment state="translated">
<source>part.table.si_value</source>
<target>SI Value</target>
</segment>
</unit>
<unit id="rW_SFJE" name="part.table.id">
<segment state="translated">
<source>part.table.id</source>