feat: migrate system + cross-DB engine + IMPORT/EXPORT syntax -- 22 files, client+server+docs
CI / test (push) Has been cancelled
CI / verify (push) Has been cancelled
Clients CI / build-server (push) Has been cancelled
Clients CI / test-python (push) Has been cancelled
Clients CI / test-javascript (push) Has been cancelled
Clients CI / test-nim (push) Has been cancelled
Clients CI / test-rust (push) Has been cancelled

This commit is contained in:
2026-05-21 19:32:14 +03:00
parent f427ba5968
commit bb843b9a03
23 changed files with 2306 additions and 52 deletions
+268
View File
@@ -0,0 +1,268 @@
# Migrations & Data Import/Export
BaraDB has a built-in migration system via BaraQL. Migrations are fully managed
by the server — checksums, locking, rollback, and status tracking. The client
sends BaraQL commands and does not maintain its own migration table.
## BaraQL Migration Syntax
```sql
-- Create a migration with UP and DOWN scripts
CREATE MIGRATION add_users_table {
UP: CREATE TABLE users (
id SERIAL PRIMARY KEY,
name VARCHAR(255) NOT NULL,
email VARCHAR(255) UNIQUE,
created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
);
DOWN: DROP TABLE IF EXISTS users;
}
-- Apply a specific migration
APPLY MIGRATION add_users_table
-- Apply all pending migrations
MIGRATION UP
-- Apply next N migrations
MIGRATION UP 3
-- Rollback last migration
MIGRATION DOWN
-- Rollback last N migrations
MIGRATION DOWN 2
-- View migration status
MIGRATION STATUS
-- Dry run (validate without executing)
MIGRATION DRY RUN add_users_table
```
## Migration Locking
BaraDB acquires a global migration lock before applying any migration. This
prevents concurrent migration runs and ensures:
- Only one migration runs at a time
- Checksums are verified before execution
- Failed migrations can be rolled back safely
- `MIGRATION STATUS` shows exact state of each migration
## Checksums
Every migration body is SHA-256 hashed. The server stores the checksum at
creation time and verifies it before applying. This prevents accidental
modification of already-registered migrations.
```
CREATE MIGRATION add_users → checksum: a3f2b8c1...
APPLY MIGRATION add_users → verifies checksum matches → executes
```
## Rollback
Migrations with `DOWN` scripts support rollback:
```sql
CREATE MIGRATION add_column {
UP: ALTER TABLE users ADD COLUMN phone VARCHAR(20);
DOWN: ALTER TABLE users DROP COLUMN phone;
}
```
When you run `MIGRATION DOWN`, the server executes the DOWN script and marks
the migration as rolled back.
## Dry Run
Validate a migration before applying:
```sql
MIGRATION DRY RUN add_users_table
-- Output:
-- DRY RUN add_users_table:
-- Statements: 1
-- [1] nkCreateTable
-- DOWN script: yes
-- Checksum: a3f2b8c1d4e5f6a7b8c9d0e1f2a3b4c5
```
---
## IMPORT FROM / EXPORT TO
BaraDB supports importing and exporting data directly via BaraQL.
### IMPORT FROM
```sql
-- Import from CSV
IMPORT FROM '/data/users.csv' INTO users
FORMAT CSV
DELIMITER ','
HEADER true
BATCH 1000
-- Import from JSON array
IMPORT FROM '/data/users.json' INTO users
FORMAT JSON
-- Import from NDJSON (newline-delimited JSON)
IMPORT FROM '/data/users.ndjson' INTO users
FORMAT NDJSON
```
Options:
| Option | Values | Default | Description |
|--------|--------|---------|-------------|
| `FORMAT` | `CSV`, `JSON`, `NDJSON` | `CSV` | Input file format |
| `DELIMITER` | any char | `,` | CSV field delimiter |
| `HEADER` | `true`/`false` | `true` | CSV first row is header |
| `BATCH` | integer | `1000` | Rows per insert batch |
### EXPORT TO
```sql
-- Export to CSV
EXPORT TO '/backup/users.csv' FROM users
FORMAT CSV
DELIMITER ','
HEADER true
-- Export to JSON
EXPORT TO '/backup/users.json' FROM users
FORMAT JSON
-- Export to NDJSON
EXPORT TO '/backup/users.ndjson' FROM users
FORMAT NDJSON
```
---
## Cross-Database Migration
BaraDB's Nim client (nim-allographer) includes a cross-database migration
engine. Migrate data from PostgreSQL, MySQL, MariaDB, SQLite, or SurrealDB
directly to BaraDB.
### Supported Sources
| Database | Schema Extraction | Status |
|----------|-------------------|--------|
| PostgreSQL | `information_schema` | ✅ |
| MySQL | `information_schema` | ✅ |
| MariaDB | `information_schema` | ✅ |
| SQLite | `sqlite_master` + `PRAGMA` | ✅ |
| SurrealDB | `INFO FOR DB` / `INFO FOR TABLE` | ✅ |
### Nim API
```nim
import allographer/migrate_data
# Connect to source and target
let pg = dbOpen(PostgreSQL, "sourcedb", "user", "pass", "localhost", 5432)
let bdb = dbOpen(Baradb, "targetdb", "admin", "", "127.0.0.1", 9472)
# Migrate all tables
let report = waitFor migrate(pg, bdb, batchSize = 5000)
echo report
# Migration: PostgreSQL → BaraDB
# Tables: 12/12
# Rows: 45230
# Time: 3.2s
# Migrate specific tables
let report = waitFor migrate(pg, bdb,
tables = @["users", "orders", "products"])
```
### Type Mapping
The migration engine automatically maps source types to BaraDB equivalents:
| PostgreSQL | MySQL | SQLite | BaraDB |
|------------|-------|--------|--------|
| `SERIAL` | `INT AUTO_INCREMENT` | `INTEGER PK` | `SERIAL` |
| `VARCHAR(n)` | `VARCHAR(n)` | `TEXT` | `VARCHAR(n)` |
| `TEXT` | `TEXT` | `TEXT` | `TEXT` |
| `BOOLEAN` | `TINYINT(1)` | `INTEGER` | `BOOLEAN` |
| `JSONB` | `JSON` | `TEXT` | `JSON` |
| `TIMESTAMP` | `DATETIME` | `TEXT` | `TIMESTAMP` |
| `UUID` | `CHAR(36)` | `TEXT` | `UUID` |
Full type map: 50+ types supported.
---
## Nim allographer Client API
### Migration Management
```nim
import allographer/query_builder/models/baradb/baradb_exec
# Create migration
let qr = waitFor rdb.createMigration("add_products",
"CREATE TABLE products (id SERIAL PRIMARY KEY, name VARCHAR(255))",
"DROP TABLE IF EXISTS products")
# Apply
let qr = waitFor rdb.applyMigration("add_products")
# Apply all pending
let qr = waitFor rdb.migrateUp()
# Rollback
let qr = waitFor rdb.migrateDown(1)
# Status
let status = waitFor rdb.migrationStatus()
# Check if applied
if waitFor rdb.isMigrationApplied("add_products"):
echo "Already applied"
# Dry run
let qr = waitFor rdb.migrationDryRun("add_products")
```
### Prepared Statements
```nim
let stmt = waitFor rdb.prepare(
"SELECT * FROM users WHERE age > ? AND status = ?", nArgs = 2)
let results = waitFor stmt.preparedGet(@[
WireValue(kind: fkInt32, int32Val: 18),
WireValue(kind: fkString, strVal: "active")
])
stmt.flushStmt()
rdb.clearStmtCache()
```
### Pagination
```nim
# Offset-based
let page = waitFor rdb.table("users").paginate(page = 1, perPage = 20)
# Cursor-based (faster for large tables)
let batch = waitFor rdb.table("users")
.fastPaginate("id", perPage = 100, afterId = "42")
```
---
## Best Practices
1. **Always include DOWN scripts** — enables safe rollback
2. **Use dry run first** — validate migrations before applying in production
3. **Batch large imports** — use `BATCH 1000` for CSV imports to avoid memory issues
4. **Export before migration** — backup data with `EXPORT TO` before cross-DB migration
5. **Check status after migration** — verify with `MIGRATION STATUS`
6. **Migrate tables without foreign keys first** — then tables with foreign keys