feat(migrate): onboarding backfill (repositories + dast_targets -> onboarded_targets) (#141)

compliance-agent/src/main.rs

@@ -1,4 +1,50 @@
-use compliance_agent::{agent, api, config, database, scheduler, ssh, webhooks};
+use compliance_agent::{agent, api, config, database, migrate, scheduler, ssh, webhooks};
+
+/// Run the `migrate onboarding` subcommand and exit. Backfills (or reverts) the
+/// unified `onboarded_targets` collection per tenant.
+///
+/// Usage: `compliance-agent migrate onboarding [--all | --tenant <id>] [--dry-run] [--revert]`
+async fn run_migration(
+    args: &[String],
+    pool: &database::DatabasePool,
+) -> Result<(), compliance_agent::error::AgentError> {
+    if args.get(2).map(String::as_str) != Some("onboarding") {
+        eprintln!(
+            "usage: compliance-agent migrate onboarding [--all | --tenant <id>] [--dry-run] [--revert]"
+        );
+        std::process::exit(2);
+    }
+    let has = |flag: &str| args.iter().any(|a| a == flag);
+    let dry_run = has("--dry-run");
+    let revert = has("--revert");
+    let tenant = args
+        .iter()
+        .position(|a| a == "--tenant")
+        .and_then(|i| args.get(i + 1))
+        .cloned();
+
+    let tenants: Vec<String> = if has("--all") {
+        pool.list_tenant_ids().await?
+    } else if let Some(t) = tenant {
+        vec![t]
+    } else {
+        eprintln!("specify --all or --tenant <id>");
+        std::process::exit(2);
+    };
+
+    for tenant_id in tenants {
+        let db = pool.for_tenant_id(&tenant_id).await?;
+        if revert {
+            migrate::onboarding::revert(&db).await?;
+            println!("[{tenant_id}] reverted onboarding backfill");
+        } else {
+            let report = migrate::onboarding::backfill_onboarded_targets(&db, dry_run).await?;
+            let prefix = if dry_run { "(dry-run) " } else { "" };
+            println!("[{tenant_id}] {prefix}{report:?}");
+        }
+    }
+    Ok(())
+}
 
 #[tokio::main]
 async fn main() -> Result<(), Box<dyn std::error::Error>> {
@@ -31,6 +77,13 @@ async fn main() -> Result<(), Box<dyn std::error::Error>> {
     let db_pool =
         database::DatabasePool::connect(&config.mongodb_uri, &config.mongodb_database).await?;
 
+    // One-shot subcommands run and exit without starting the servers.
+    let args: Vec<String> = std::env::args().collect();
+    if args.get(1).map(String::as_str) == Some("migrate") {
+        run_migration(&args, &db_pool).await?;
+        return Ok(());
+    }
+
     let agent = agent::ComplianceAgent::new(config.clone(), db_pool);
 
     tracing::info!("Starting scheduler...");