sad_spirit / pg_builder
Query builder for Postgres backed by SQL parser
Installs: 23 433
Dependents: 1
Suggesters: 0
Security: 0
Stars: 56
Watchers: 2
Forks: 6
Open Issues: 0
Requires
- php: >=7.2.0
- ext-ctype: *
Requires (Dev)
- ext-pdo_pgsql: *
- ext-pgsql: *
- phpstan/phpstan: ^1.10.15
- phpunit/phpunit: ^8.0|^9.0
- psr/cache: ^1.0
- sad_spirit/pg_wrapper: ^2.4.0
- vimeo/psalm: ^4.30
Suggests
- ext-pdo_pgsql: Used for running built queries
- psr/cache-implementation: Used for caching parsed queries
- sad_spirit/pg_wrapper: Used for DB types conversion and running built queries
Conflicts
- sad_spirit/pg_wrapper: <2.4
This package is auto-updated.
Last update: 2024-10-27 11:34:17 UTC
README
This is a query builder for Postgres with a twist: it contains a partial1 reimplementation of PostgreSQL's own query parser. This sets it aside from the usual breed of "write-only" query builders:
- Query is represented as an Abstract Syntax Tree quite similar to PostgreSQL's internal representation.
- Query parts can be added to the AST either as objects or as strings (that will be processed by Parser).
- Nodes can be removed and replaced in AST.
- AST can be analyzed and transformed, the package takes advantage of this to allow named parameters like
:foo
instead of standard PostgreSQL's positional parameters$1
and to infer parameters' types from SQL typecasts. - Almost all syntax available for
SELECT
(andVALUES
) /INSERT
/UPDATE
/DELETE
/MERGE
in PostgreSQL 16 is supported, query being built is automatically checked for correct syntax.
Substantial effort was made to optimise parsing, but not parsing is faster anyway, so there are means to cache parts of AST and the resultant query.
Usage example
use sad_spirit\pg_builder\{ Select, StatementFactory, converters\BuilderSupportDecorator }; use sad_spirit\pg_wrapper\{ Connection, converters\DefaultTypeConverterFactory }; $wantPDO = false; if ($wantPDO) { $pdo = new \PDO('pgsql:host=localhost;user=username;dbname=cms'); // Uses DB connection properties to set up parsing and building of SQL $factory = StatementFactory::forPDO($pdo); // NB: This still requires sad_spirit/pg_wrapper for type conversion code $converter = new BuilderSupportDecorator(new DefaultTypeConverterFactory(), $factory->getParser()); } else { $connection = new Connection('host=localhost user=username dbname=cms'); // Uses DB connection properties to set up parsing and building of SQL $factory = StatementFactory::forConnection($connection); // Needed for handling type info extracted from query $connection->setTypeConverterFactory(new BuilderSupportDecorator( $connection->getTypeConverterFactory(), $factory->getParser() )); } // latest 5 news /** @var Select $query */ $query = $factory->createFromString( 'select n.* from news as n order by news_added desc limit 5' ); // we also need pictures for these... $query->list[] = 'p.*'; $query->from[0]->leftJoin('pictures as p')->on = 'n.picture_id = p.picture_id'; // ...and need to limit them to only specific rubrics $query->from[] = 'objects_rubrics as ro'; $query->where->and('ro.rubric_id = any(:rubric::integer[]) and ro.obj_id = n.news_id'); // ...and keep 'em fresh $query->where->and('age(news_added) < :age::interval'); // $generated contains a query, mapping from named parameters to positional ones, types info // it can be easily cached to prevent parsing/building SQL on each request $generated = $factory->createFromAST($query); // Note that we don't have to specify parameter types, these are extracted from query if ($wantPDO) { $result = $pdo->prepare($generated->getSql()); $result->execute($converter->convertParameters( $generated, [ 'rubric' => [19, 20, 21], 'age' => 30 * 24 * 3600 ] )); } else { $result = $generated->executeParams( $connection, [ 'rubric' => [19, 20, 21], 'age' => 30 * 24 * 3600 ] ); } foreach ($result as $row) { print_r($row); } echo $generated->getSql();
the last echo
statement will output something like
select n.*, p.* from news as n left join pictures as p on n.picture_id = p.picture_id, objects_rubrics as ro where ro.rubric_id = any($1::pg_catalog.int4[]) and ro.obj_id = n.news_id and age(news_added) < $2::interval order by news_added desc limit 5
if targeting Connection
and something like
select n.*, p.* from news as n left join pictures as p on n.picture_id = p.picture_id, objects_rubrics as ro where ro.rubric_id = any(:rubric::pg_catalog.int4[]) and ro.obj_id = n.news_id and age(news_added) < :age::interval order by news_added desc limit 5
if targeting PDO
Documentation
Is in the wiki
1: "Partial" here means the following: PostgreSQL grammar file src/backend/parser/gram.y
is about 19K lines long.
Of these about 5K lines are used for SELECT
/ INSERT
/ UPDATE
/ DELETE
/ MERGE
queries and are reimplemented here.