Janady13/missionmatch / internal/platform/database/database.go

Source identity

Field	Recorded value
Source type	GitHub source
Original record	Janady13/missionmatch / internal/platform/database/database.go
Date	Not specified in source
Repository	Janady13/missionmatch
Revision	39df94f92b2a33f852481fe1954dabed841524bb
Branch	w0-aedt-compliance
Path or archive member	internal/platform/database/database.go
Original SHA-256	58e4880187ac752977378dd77c8bf4a1e956e01ff2aa7316858cd379d24f4b86
Git blob SHA	fc8e07db8cb13c35271c2a247f962dede7337a17
Reading-copy SHA-256	58e4880187ac752977378dd77c8bf4a1e956e01ff2aa7316858cd379d24f4b86

Original source ↗

Recovered source transcript

Download complete source text ↓67 lines · 2.7 KB

// Package database constructs the application's Postgres connection pool.
//
// Two things are wired here that the rest of the app relies on (see docs/STACK_DECISIONS.md §2/§3/§5):
//
//  1. pgvector codecs are registered on EVERY physical connection via AfterConnect, so vector
//     columns bind and scan as pgvector.Vector. Registering once on a single conn is a bug under
//     pool churn — it must be AfterConnect.
//  2. The pool is expected to authenticate as the NOBYPASSRLS application role (never the migrator),
//     so Postgres Row-Level Security is always in force. Which role is used is a property of the
//     DSN, not of this code.
package database

import (
	"context"
	"fmt"

	"github.com/jackc/pgx/v5"
	"github.com/jackc/pgx/v5/pgxpool"
	pgxvec "github.com/pgvector/pgvector-go/pgx"
)

// NewPool parses dsn, registers the pgvector codecs per connection, and returns a ready pool.
// The caller owns the returned pool's Close(). dsn MUST point at the NOBYPASSRLS application role.
func NewPool(ctx context.Context, dsn string) (*pgxpool.Pool, error) {
	cfg, err := pgxpool.ParseConfig(dsn)
	if err != nil {
		return nil, fmt.Errorf("parse pool config: %w", err)
	}

	// AfterConnect runs once per physical connection the pool opens — the only correct place to
	// register vector/halfvec/sparsevec so pgvector.Vector works as a bind param and Scan target.
	cfg.AfterConnect = func(ctx context.Context, conn *pgx.Conn) error {
		if err := pgxvec.RegisterTypes(ctx, conn); err != nil {
			return fmt.Errorf("register pgvector types: %w", err)
		}
		return nil
	}

	pool, err := pgxpool.NewWithConfig(ctx, cfg)
	if err != nil {
		return nil, fmt.Errorf("create pool: %w", err)
	}
	return pool, nil
}

// AssertRole fails closed unless the pool authenticates as wantUser with rolbypassrls=false. It is
// the runtime guardrail behind the whole RLS and EEO isolation story: if an operator pastes a
// BYPASSRLS or superuser DSN (for example the migrator DSN) into the app or compliance variable, RLS
// would be silently bypassed and every tenant's rows, including protected demographic attributes,
// would leak cross org. This converts that misconfiguration into a hard boot failure.
func AssertRole(ctx context.Context, pool *pgxpool.Pool, wantUser string) error {
	var user string
	var bypass bool
	err := pool.QueryRow(ctx,
		`SELECT current_user, rolbypassrls FROM pg_roles WHERE rolname = current_user`).Scan(&user, &bypass)
	if err != nil {
		return fmt.Errorf("role assertion query: %w", err)
	}
	if user != wantUser {
		return fmt.Errorf("pool authenticates as %q, expected %q", user, wantUser)
	}
	if bypass {
		return fmt.Errorf("pool role %q has BYPASSRLS, which would bypass row level security", user)
	}
	return nil
}