export class DataMigrationService {
source;
target;
constructor(source, target) {
this.source = source;
this.target = target;
}
async migrateTable(tableName, targetTableName, batchSize = 1000) {
const start = Date.now();
const targetName = targetTableName || tableName;
// 1. Describe source to know columns
const columns = await this.source.describeTable(tableName);
const colNames = columns.map(c => c.name);
// 2. Read data (Simulated stream via simple query for MVP)
// In production, this should be a true stream/cursor.
// Here we fetch up to 10000 rows.
const sql = `SELECT ${colNames.join(', ')} FROM ${tableName}`;
// Using a large limit for MVP migration
const result = await this.source.executeQuery(sql, 100000);
const rows = result.rows;
if (rows.length === 0) {
return { rows: 0, time: Date.now() - start };
}
// 3. Insert in batches
let totalInserted = 0;
for (let i = 0; i < rows.length; i += batchSize) {
const batch = rows.slice(i, i + batchSize);
// Basic Type Transformation (Mapping dates etc if needed)
// Currently insertBatch in DatabaseService handles basic insertion logic
await this.target.insertBatch(targetName, batch);
totalInserted += batch.length;
}
return {
rows: totalInserted,
time: Date.now() - start
};
}
async migrateData(tables, truncateTarget = false) {
const results = [];
for (const table of tables) {
if (truncateTarget) {
// TODO: Implement truncate in DatabaseService
// await this.target.truncateTable(table);
}
try {
const res = await this.migrateTable(table);
results.push({ table, status: 'success', rows: res.rows, time: res.time });
}
catch (e) {
results.push({ table, status: 'error', error: e.message });
}
}
return results;
}
}
