feat(iceberg): storage & catalog parity for the ingestr destination

[iremcaginyurtturk]

Follow-up to #2383 (which added Iceberg as an ingestr destination). This closes the remaining gaps so the Bruin connection can express ingestr's full Iceberg URI surface.

Catalog

• REST: rest_use_ssl to select HTTPS — required for managed catalogs (Polaris, Unity, Lakekeeper, Tabular, R2); ingestr defaults REST to plain HTTP.
• SQL: driver / dialect for the generic sql catalog (ingestr requires both; the dedicated sqlite/postgres types set them automatically).

Storage

• Native GCS (type: gcs): gs:// warehouse with key_file / key_json service-account credentials (or ADC).
• Local filesystem (type: local): a filesystem path → file:// warehouse.
• Hadoop object-storage: path is now optional, so the warehouse can come from the storage block (s3://, gs://) instead of a local dir.
• Storage block is optional when the catalog supplies its own warehouse (Glue/REST/SQL).

Docs

• Per-backend storage examples: AWS S3, S3 bucket+prefix, S3-compatible (MinIO/R2), native GCS, GCS via the S3-interop endpoint (HMAC), and local.
• Postgres catalog DSN passthrough (sslmode, sslrootcert, connect_timeout, …) via properties, scoped to type: postgres.

[greptile-apps]

Prompt To Fix All With AI

### Issue 1
pkg/iceberg/config.go:217-220
**Warehouse-less catalogs pass validation**

When SQLite, Postgres, or Hive omits `storage`, or Hadoop has neither `catalog.path` nor `storage`, this branch emits an Iceberg URI without a warehouse, causing ingestr destination initialization or table creation to fail at runtime.

---

For each issue above, determine whether it is valid and should be fixed. If so, fix it directly.

_{Reviews (1): Last reviewed commit: "feat(iceberg): storage & catalog parity ..." | Re-trigger Greptile}

[greptile-apps]

_{Reviews (2): Last reviewed commit: "feat(iceberg): storage & catalog parity ..." | Re-trigger Greptile}

[greptile-apps]

Prompt To Fix All With AI

### Issue 1
pkg/iceberg/config.go:100-102
**Catalog-owned SQL warehouse rejected**

When a generic SQL catalog supplies its own warehouse and the connection omits storage, `catalogNeedsWarehouse` still requires a client-supplied warehouse, causing `GetIngestrURI` to reject the valid connection before ingestr starts.

```suggestion
	case config.IcebergCatalogSQLite, config.IcebergCatalogPostgres,
		config.IcebergCatalogHive, config.IcebergCatalogHadoop:
		return true
```

---

For each issue above, determine whether it is valid and should be fixed. If so, fix it directly.

_{Reviews (3): Last reviewed commit: "feat(iceberg): storage & catalog parity ..." | Re-trigger Greptile}

[greptile-apps]

_{Reviews (4): Last reviewed commit: "feat(iceberg): storage & catalog parity ..." | Re-trigger Greptile}

[greptile-apps]

Prompt To Fix All With AI

### Issue 1
pkg/iceberg/config.go:214-217
**Optional storage remains schema-required**

When a Glue, REST, or SQL catalog supplies its warehouse and omits `storage`, this branch accepts the configuration, but the reflected connection schema still requires `storage`, causing schema-aware clients to reject the newly supported configuration before `GetIngestrURI` runs.

---

For each issue above, determine whether it is valid and should be fixed. If so, fix it directly.

_{Reviews (5): Last reviewed commit: "feat(iceberg): storage & catalog parity ..." | Re-trigger Greptile}

[greptile-apps]

_{Reviews (6): Last reviewed commit: "feat(iceberg): storage & catalog parity ..." | Re-trigger Greptile}

[greptile-apps]

_{Reviews (7): Last reviewed commit: "feat(iceberg): storage & catalog parity ..." | Re-trigger Greptile}

[greptile-apps]

_{Reviews (8): Last reviewed commit: "feat(iceberg): storage & catalog parity ..." | Re-trigger Greptile}