re_redap_tests 0.34.0

#![expect(clippy::unwrap_used)]

use std::sync::Arc;

use arrow::array::{Float32Array, Float64Array, ListArray, RecordBatch, StringArray};
use arrow::datatypes::Schema;
use futures::TryStreamExt as _;
use itertools::{Itertools as _, izip};
use re_arrow_util::ArrowArrayDowncastRef as _;
use re_log_types::{EntityPath, TimeType};
use re_protos::cloud::v1alpha1::ext::{
    self, DatasetDetails, RegisterWithDatasetRequest, TableDetails, UpdateTableEntryRequest,
};
use re_protos::cloud::v1alpha1::ext::{ScanDatasetManifestDataframe, ScanSegmentTableDataframe};
use re_protos::cloud::v1alpha1::rerun_cloud_service_server::RerunCloudService;
use re_protos::cloud::v1alpha1::{
    CreateDatasetEntryRequest, GetDatasetManifestSchemaRequest, GetSegmentTableSchemaRequest,
    ReadDatasetEntryRequest, ReadTableEntryRequest, ScanDatasetManifestRequest,
    ScanSegmentTableRequest,
};
use re_protos::common::v1alpha1::ext::IfDuplicateBehavior;
use re_protos::headers::RerunHeadersInjectorExt as _;
use re_protos::{cloud::v1alpha1::ext as cloud_ext, common::v1alpha1::TaskId};
use re_sdk_types::{AnyValues, LayerName};
use re_types_core::AsComponents;
use re_types_core::SegmentId;
use url::Url;

use super::common::{
    DataSourcesDefinition, LayerDefinition, RerunCloudServiceExt as _,
    create_table_entry_with_name, entry_name, prop,
};
use crate::{
    FieldsTestExt as _, RecordBatchTestExt as _, SchemaTestExt as _, TuidPrefix,
    create_simple_recording_in,
};

pub async fn register_and_scan_simple_dataset(service: impl RerunCloudService) {
    let data_sources_def = DataSourcesDefinition::new_with_tuid_prefix(
        1,
        [
            LayerDefinition::simple("my_segment_id1", &["my/entity", "my/other/entity"]),
            LayerDefinition::simple("my_segment_id2", &["my/entity"]),
            LayerDefinition::simple(
                "my_segment_id3",
                &["my/entity", "another/one", "yet/another/one"],
            ),
        ],
    );

    let dataset_name = "my_dataset1";
    service.create_dataset_entry_with_name(dataset_name).await;
    service
        .register_with_dataset_name_blocking(dataset_name, data_sources_def.to_data_sources())
        .await;

    scan_segment_table_and_snapshot(&service, dataset_name, "simple").await;
    scan_dataset_manifest_and_snapshot(&service, dataset_name, "simple").await;
}

/// Make sure that registering to blueprint dataset works as expected for tables
pub async fn register_and_attach_table_blueprint_dataset(service: impl RerunCloudService) {
    let blueprint_segment_name = "table_blueprint_segment_id";
    let blueprint_segment = SegmentId::from(blueprint_segment_name);
    let table_dir = tempfile::tempdir().expect("create temp dir");
    let table =
        create_table_entry_with_name(&service, "table_with_registered_blueprint", &table_dir).await;
    let table_id = table.details.id;
    let blueprint_dataset = table
        .table_details
        .blueprint_dataset
        .expect("tables should get an implicit blueprint dataset");

    let blueprint_dataset_name =
        register_simple_blueprint_segment(&service, blueprint_dataset, blueprint_segment_name, 3)
            .await;

    let updated_table = service
        .update_table_entry(tonic::Request::new(
            UpdateTableEntryRequest {
                id: table_id,
                table_details: TableDetails {
                    blueprint_dataset: Some(blueprint_dataset),
                    default_blueprint_segment: Some(blueprint_segment.clone()),
                },
            }
            .into(),
        ))
        .await
        .expect("update table blueprint details")
        .into_inner()
        .table
        .expect("table missing in update_table response")
        .table_details
        .expect("table details");
    assert_eq!(
        updated_table.blueprint_dataset,
        Some(blueprint_dataset.into())
    );
    assert_eq!(
        updated_table.default_blueprint_segment,
        Some(blueprint_segment.clone().into())
    );

    let read_back = service
        .read_table_entry(tonic::Request::new(ReadTableEntryRequest {
            id: Some(table_id.into()),
        }))
        .await
        .expect("read table entry")
        .into_inner()
        .table
        .expect("table missing in read_table response")
        .table_details
        .expect("table details");
    assert_eq!(read_back.blueprint_dataset, Some(blueprint_dataset.into()));
    assert_eq!(
        read_back.default_blueprint_segment,
        Some(blueprint_segment.clone().into())
    );

    let segment_table = scan_segment_table(&service, &blueprint_dataset_name).await;
    assert_eq!(segment_table.num_rows(), 1);
    let manifest = scan_dataset_manifest(&service, &blueprint_dataset_name).await;
    assert_eq!(manifest.num_rows(), 1);
}

/// Make sure that registering to blueprint dataset works as expected.
pub async fn register_and_scan_blueprint_dataset(service: impl RerunCloudService) {
    let dataset_name = "my_dataset1";
    service.create_dataset_entry_with_name(dataset_name).await;

    let dataset_details: DatasetDetails = service
        .read_dataset_entry(
            tonic::Request::new(ReadDatasetEntryRequest {})
                .with_entry_name(entry_name(dataset_name)),
        )
        .await
        .unwrap()
        .into_inner()
        .dataset
        .unwrap()
        .dataset_details
        .unwrap()
        .try_into()
        .unwrap();

    assert!(dataset_details.blueprint_dataset.is_some());

    let blueprint_dataset_name = register_simple_blueprint_segment(
        &service,
        dataset_details.blueprint_dataset.unwrap(),
        "blueprint_segment_id",
        2,
    )
    .await;

    scan_segment_table_and_snapshot(&service, &blueprint_dataset_name, "simple_blueprint").await;
    scan_dataset_manifest_and_snapshot(&service, &blueprint_dataset_name, "simple_blueprint").await;
}

async fn register_simple_blueprint_segment(
    service: &impl RerunCloudService,
    blueprint_dataset: re_log_types::EntryId,
    blueprint_segment_name: &'static str,
    tuid_prefix: TuidPrefix,
) -> String {
    let blueprint_data_sources_def = DataSourcesDefinition::new_with_tuid_prefix(
        tuid_prefix,
        [LayerDefinition::simple_blueprint(blueprint_segment_name)],
    );

    let blueprint_dataset_name = service
        .read_dataset_entry(
            tonic::Request::new(ReadDatasetEntryRequest {}).with_entry_id(blueprint_dataset),
        )
        .await
        .unwrap()
        .into_inner()
        .dataset
        .unwrap()
        .details
        .unwrap()
        .name
        .unwrap();

    service
        .register_with_dataset_name_blocking(
            &blueprint_dataset_name,
            blueprint_data_sources_def.to_data_sources(),
        )
        .await;

    blueprint_dataset_name
}

pub async fn register_and_scan_simple_dataset_with_properties(service: impl RerunCloudService) {
    let data_sources_def = DataSourcesDefinition::new_with_tuid_prefix(
        1,
        [
            LayerDefinition::simple("my_segment_id1", &["my/entity", "my/other/entity"]),
            LayerDefinition::simple("my_segment_id2", &["my/entity"]),
            LayerDefinition::simple(
                "my_segment_id3",
                &["my/entity", "another/one", "yet/another/one"],
            ),
            LayerDefinition::properties(
                "my_segment_id1",
                [prop(
                    "text_log",
                    re_sdk_types::archetypes::TextLog::new("i'm segment 1"),
                )],
            )
            .layer_name("props"),
            LayerDefinition::properties(
                "my_segment_id2",
                [
                    prop(
                        "text_log",
                        re_sdk_types::archetypes::TextLog::new("i'm segment 2"),
                    ),
                    prop(
                        "points",
                        re_sdk_types::archetypes::Points2D::new([(0.0, 1.0)]),
                    ),
                ],
            )
            .layer_name("props"),
        ],
    );

    let dataset_name = "my_dataset1";
    service.create_dataset_entry_with_name(dataset_name).await;
    service
        .register_with_dataset_name_blocking(dataset_name, data_sources_def.to_data_sources())
        .await;

    scan_segment_table_and_snapshot(&service, dataset_name, "simple_with_properties").await;
    scan_dataset_manifest_and_snapshot(&service, dataset_name, "simple_with_properties").await;
}

/// This test checks that the registration order takes precedence to resolve a segment's
/// properties.
///
/// Note: this is not great. We should probably use the "regular" Rerun way for that (aka row id
/// timestamp). But this is not how Rerun Hub is currently working, and consistency is better than
/// correctness for the OSS server.
pub async fn register_and_scan_simple_dataset_with_properties_out_of_order(
    service: impl RerunCloudService,
) {
    let first_logged_data_sources_def = DataSourcesDefinition::new_with_tuid_prefix(
        10, // <- mind this
        [LayerDefinition::properties(
            "my_segment_id1",
            [prop(
                "text_log",
                re_sdk_types::archetypes::TextLog::new(
                    "I was logged first, registered last, so I should win",
                ),
            )],
        )
        .layer_name("prop1")],
    );
    let first_logged_data_sources = first_logged_data_sources_def.to_data_sources();

    let last_logged_data_sources_def = DataSourcesDefinition::new_with_tuid_prefix(
        20, // <- mind this
        [LayerDefinition::properties(
            "my_segment_id1",
            [prop(
                "text_log",
                re_sdk_types::archetypes::TextLog::new("I was logged last, registered first"),
            )],
        )
        .layer_name("prop2")],
    );
    let last_logged_data_sources = last_logged_data_sources_def.to_data_sources();

    let dataset_name = "my_dataset";
    service.create_dataset_entry_with_name(dataset_name).await;
    service
        .register_with_dataset_name_blocking(dataset_name, last_logged_data_sources)
        .await;

    service
        .register_with_dataset_name_blocking(dataset_name, first_logged_data_sources)
        .await;

    let dataset_manifest =
        scan_dataset_manifest_and_snapshot(&service, dataset_name, "out_of_order_properties").await;
    scan_segment_table_and_snapshot(&service, dataset_name, "out_of_order_properties").await;

    // assert test correctness
    let registration_time_col = ScanDatasetManifestDataframe::COLUMN_RERUN_REGISTRATION_TIME
        .extract(&dataset_manifest)
        .unwrap();

    let prop_col = dataset_manifest
        .column_by_name("property:text_log:TextLog:text")
        .unwrap()
        .downcast_array_ref::<ListArray>()
        .unwrap();

    assert!(registration_time_col[0] < registration_time_col[1]);

    assert_eq!(
        prop_col
            .value(0)
            .downcast_array_ref::<StringArray>()
            .unwrap()
            .value(0),
        "I was logged last, registered first"
    );
}

pub async fn register_and_scan_simple_dataset_with_layers(service: impl RerunCloudService) {
    let data_sources_def = DataSourcesDefinition::new_with_tuid_prefix(
        1,
        [
            LayerDefinition::simple("segment1", &["my/entity", "another/one", "yet/another/one"]),
            LayerDefinition::simple("segment1", &["extra/entity"]).layer_name("extra"),
            LayerDefinition::simple("segment2", &["another/one", "yet/another/one"])
                .layer_name("base"),
            LayerDefinition::simple("segment2", &["extra/entity"]).layer_name("extra"),
            LayerDefinition::simple("segment3", &["i/am/alone"]),
        ],
    );

    let dataset_name = "dataset_with_layers";
    service.create_dataset_entry_with_name(dataset_name).await;
    service
        .register_with_dataset_name_blocking(dataset_name, data_sources_def.to_data_sources())
        .await;

    scan_segment_table_and_snapshot(&service, dataset_name, "simple_with_layers").await;
    scan_dataset_manifest_and_snapshot(&service, dataset_name, "simple_with_layers").await;
}

pub async fn register_and_scan_simple_dataset_multiple_timelines(service: impl RerunCloudService) {
    let data_sources_def = DataSourcesDefinition::new_with_tuid_prefix(
        1,
        [
            LayerDefinition::simple_with_time(
                "my_segment_id1",
                &["my/entity", "my/other/entity"],
                0,
                TimeType::Sequence,
            ),
            LayerDefinition::simple_with_time(
                "my_segment_id2",
                &["my/entity"],
                0,
                TimeType::DurationNs,
            ),
            LayerDefinition::properties(
                "my_segment_id1",
                [prop(
                    "text_log",
                    re_sdk_types::archetypes::TextLog::new("i'm segment 1"),
                )],
            )
            .layer_name("props"),
            LayerDefinition::simple_with_time(
                "my_segment_id3",
                &["my/entity", "another/one", "yet/another/one"],
                0,
                TimeType::TimestampNs,
            ),
            LayerDefinition::simple_with_time(
                "my_segment_id2",
                &["my/entity", "my/fourth/entity"],
                0,
                TimeType::Sequence,
            )
            .layer_name("layer_two"),
        ],
    );

    let dataset_name = "my_dataset1";
    service.create_dataset_entry_with_name(dataset_name).await;
    service
        .register_with_dataset_name_blocking(dataset_name, data_sources_def.to_data_sources())
        .await;

    scan_segment_table_and_snapshot(&service, dataset_name, "simple_with_multiple_timelines").await;
    scan_dataset_manifest_and_snapshot(&service, dataset_name, "simple_with_multiple_timelines")
        .await;
}

pub async fn register_with_prefix(fe: impl RerunCloudService) {
    let root_dir = tempfile::tempdir().expect("creating temp dir");

    // Note: for this test, we don't use teh `DataSourceDefinition` abstraction here because we need
    // tight control of where the RRDs are stored.
    let tuid_prefix1 = 1;
    create_simple_recording_in(
        tuid_prefix1,
        "my_segment_id1",
        &["my/entity", "my/other/entity"],
        0,
        TimeType::Sequence,
        root_dir.path(),
    )
    .expect("creating recording");

    let tuid_prefix2 = 2;
    create_simple_recording_in(
        tuid_prefix2,
        "my_segment_id2",
        &["my/entity"],
        0,
        TimeType::Sequence,
        root_dir.path(),
    )
    .expect("creating recording");

    let tuid_prefix3 = 3;
    create_simple_recording_in(
        tuid_prefix3,
        "my_segment_id3",
        &["my/entity", "another/one", "yet/another/one"],
        0,
        TimeType::Sequence,
        root_dir.path(),
    )
    .expect("creating recording");

    let dataset_name = "my_dataset1";
    fe.create_dataset_entry(tonic::Request::new(CreateDatasetEntryRequest {
        name: Some(dataset_name.to_owned()),
        id: None,
    }))
    .await
    .unwrap();

    let root_url =
        Url::parse(&format!("file://{}/", root_dir.path().display())).expect("creating root url");

    fe.register_with_dataset_name_blocking(
        dataset_name,
        vec![cloud_ext::DataSource::new_rrd_prefix_url(root_url).into()],
    )
    .await;

    scan_segment_table_and_snapshot(&fe, dataset_name, "register_prefix_segments").await;
    scan_dataset_manifest_and_snapshot(&fe, dataset_name, "register_prefix_manifest").await;
}

// Scanning an empty dataset should return an empty dataframe with the expected schema -- not a
// NOT_FOUND error.
pub async fn register_and_scan_empty_dataset(service: impl RerunCloudService) {
    let dataset_name = "empty_dataset";
    service.create_dataset_entry_with_name(dataset_name).await;

    scan_segment_table_and_snapshot(&service, dataset_name, "empty").await;
    scan_dataset_manifest_and_snapshot(&service, dataset_name, "empty").await;
}

/// Any kind of bad file URI should return a not found error.
///
/// This includes:
/// - file not found
/// - path is not a file
/// - URI has a host name
///
/// The latter can be caused by attempting to build a `file://` with a relative path, leading to
/// `file://path/to/file.rrd`. This is valid URI, but here `path` is the hostname.
pub async fn register_bad_file_uri_should_error(service: impl RerunCloudService) {
    let temp_dir = tempfile::tempdir().expect("creating temp dir");
    let temp_dir_uri = format!("file://{}/", temp_dir.path().display());

    let test_cases = vec![
        ("file doesn't exist", "file:///does/not/exist.rrd"),
        ("URI has a host name", "file://somehost/file/path.rrd"),
        ("URI points to a directory", &temp_dir_uri),
    ];

    let dataset_name = "empty_dataset";
    service.create_dataset_entry_with_name(dataset_name).await;

    for (test_name, bad_uri) in test_cases {
        let request = RegisterWithDatasetRequest {
            data_sources: vec![cloud_ext::DataSource::new_rrd_url(
                url::Url::parse(bad_uri).unwrap(),
            )],
            on_duplicate: Default::default(),
        };

        let result = service
            .register_with_dataset(
                tonic::Request::new(request.into()).with_entry_name(entry_name(dataset_name)),
            )
            .await;

        assert!(
            result.is_err(),
            "register on unknown file should fail (case: {test_name})"
        );
        assert_eq!(
            result.unwrap_err().code(),
            tonic::Code::NotFound,
            "bad file URI should result in a not found error (case: {test_name})"
        );
    }
}

pub async fn register_segment_bumps_timestamp(service: impl RerunCloudService) {
    async fn get_dataset_updated_at_nanos(
        service: &impl RerunCloudService,
        dataset_name: &str,
    ) -> i64 {
        service
            .read_dataset_entry(
                tonic::Request::new(ReadDatasetEntryRequest {})
                    .with_entry_name(entry_name(dataset_name)),
            )
            .await
            .unwrap()
            .into_inner()
            .dataset
            .unwrap()
            .details
            .as_ref()
            .unwrap()
            .updated_at
            .as_ref()
            .map(|ts| ts.seconds * 1_000_000_000 + ts.nanos as i64)
            .unwrap()
    }

    let dataset_name = "timestamp_test_dataset";

    //
    // Create a dataset
    //

    service.create_dataset_entry_with_name(dataset_name).await;

    let initial_updated_at_nanos = get_dataset_updated_at_nanos(&service, dataset_name).await;

    // Small delay to ensure timestamp difference
    tokio::time::sleep(tokio::time::Duration::from_millis(10)).await;

    //
    // Register a segment - this should update the timestamp
    //

    let data_sources_def = DataSourcesDefinition::new_with_tuid_prefix(
        1,
        [LayerDefinition::simple("segment1", &["my/entity"])],
    );

    service
        .register_with_dataset_name_blocking(dataset_name, data_sources_def.to_data_sources())
        .await;

    let after_register_updated_at_nanos =
        get_dataset_updated_at_nanos(&service, dataset_name).await;

    assert!(
        after_register_updated_at_nanos > initial_updated_at_nanos,
        "Timestamp should be updated after registering segment. Initial: {initial_updated_at_nanos}, After register: {after_register_updated_at_nanos}"
    );

    // Small delay to ensure timestamp difference
    tokio::time::sleep(tokio::time::Duration::from_millis(10)).await;

    //
    // Register another layer to the same segment - this should also update the timestamp
    //

    let layer_data_sources_def = DataSourcesDefinition::new_with_tuid_prefix(
        2,
        [LayerDefinition::simple("segment1", &["another/entity"]).layer_name("layer2")],
    );

    service
        .register_with_dataset_name_blocking(dataset_name, layer_data_sources_def.to_data_sources())
        .await;

    let after_layer_updated_at_nanos = get_dataset_updated_at_nanos(&service, dataset_name).await;

    assert!(
        after_layer_updated_at_nanos > after_register_updated_at_nanos,
        "Timestamp should be updated after adding a layer. After register: {after_register_updated_at_nanos}, After layer: {after_layer_updated_at_nanos}"
    );
}

/// Tests that registering the same segment twice with `IfDuplicateBehavior::Error` fails.
pub async fn register_with_dataset_if_duplicate_behavior_error(service: impl RerunCloudService) {
    let dataset_name = "duplicate_error_test";
    service.create_dataset_entry_with_name(dataset_name).await;

    // First registration - should succeed
    let data_sources_def = DataSourcesDefinition::new_with_tuid_prefix(
        1,
        [LayerDefinition::simple("segment1", &["my/entity"])],
    );

    service
        .register_with_dataset_name_blocking_with_behavior(
            dataset_name,
            data_sources_def.to_data_sources(),
            IfDuplicateBehavior::Error,
        )
        .await;

    // Second registration of same segment - should fail with AlreadyExists
    let request = cloud_ext::RegisterWithDatasetRequest {
        data_sources: data_sources_def.to_data_sources_ext(),
        on_duplicate: IfDuplicateBehavior::Error,
    };

    let result = service
        .register_with_dataset(
            tonic::Request::new(request.into()).with_entry_name(entry_name(dataset_name)),
        )
        .await;

    assert!(
        result.is_err(),
        "second registration with Error behavior should fail"
    );
    assert_eq!(
        result.unwrap_err().code(),
        tonic::Code::AlreadyExists,
        "second registration should return AlreadyExists error"
    );
}

/// Tests that registering a duplicate with `IfDuplicateBehavior::Skip` succeeds but keeps original data.
pub async fn register_with_dataset_if_duplicate_behavior_skip(service: impl RerunCloudService) {
    let dataset_name = "duplicate_skip_test";
    service.create_dataset_entry_with_name(dataset_name).await;

    // First registration with properties
    let first_data_sources_def = DataSourcesDefinition::new_with_tuid_prefix(
        1,
        [LayerDefinition::properties(
            "segment1",
            [prop(
                "text_log",
                re_sdk_types::archetypes::TextLog::new("first"),
            )],
        )],
    );

    service
        .register_with_dataset_name_blocking_with_behavior(
            dataset_name,
            first_data_sources_def.to_data_sources(),
            IfDuplicateBehavior::Skip,
        )
        .await;

    // Second registration with different properties - should be skipped
    let second_data_sources_def = DataSourcesDefinition::new_with_tuid_prefix(
        2,
        [LayerDefinition::properties(
            "segment1",
            [prop(
                "text_log",
                re_sdk_types::archetypes::TextLog::new("second"),
            )],
        )],
    );

    service
        .register_with_dataset_name_blocking_with_behavior(
            dataset_name,
            second_data_sources_def.to_data_sources(),
            IfDuplicateBehavior::Skip,
        )
        .await;

    // Verify the property value is still "first"
    let dataset_manifest = scan_dataset_manifest(&service, dataset_name).await;

    let prop_col = dataset_manifest
        .column_by_name("property:text_log:TextLog:text")
        .expect("property column should exist")
        .downcast_array_ref::<ListArray>()
        .expect("property column should be a list array");

    let inner_array = prop_col.value(0);
    let string_array = inner_array
        .downcast_array_ref::<StringArray>()
        .expect("inner array should be string array");
    let text = string_array.value(0);

    assert_eq!(
        text, "first",
        "property should still be 'first' after Skip behavior"
    );
}

/// Tests that registering a duplicate with `IfDuplicateBehavior::Overwrite` replaces existing data.
pub async fn register_with_dataset_if_duplicate_behavior_overwrite(
    service: impl RerunCloudService,
) {
    let dataset_name = "duplicate_overwrite_test";
    service.create_dataset_entry_with_name(dataset_name).await;

    // First registration with properties
    let first_data_sources_def = DataSourcesDefinition::new_with_tuid_prefix(
        1,
        [LayerDefinition::properties(
            "segment1",
            [prop(
                "text_log",
                re_sdk_types::archetypes::TextLog::new("first"),
            )],
        )],
    );

    service
        .register_with_dataset_name_blocking_with_behavior(
            dataset_name,
            first_data_sources_def.to_data_sources(),
            IfDuplicateBehavior::Overwrite,
        )
        .await;

    // Second registration with different properties - should overwrite
    let second_data_sources_def = DataSourcesDefinition::new_with_tuid_prefix(
        2,
        [LayerDefinition::properties(
            "segment1",
            [prop(
                "text_log",
                re_sdk_types::archetypes::TextLog::new("second"),
            )],
        )],
    );

    service
        .register_with_dataset_name_blocking_with_behavior(
            dataset_name,
            second_data_sources_def.to_data_sources(),
            IfDuplicateBehavior::Overwrite,
        )
        .await;

    // Verify the property value is now "second"
    let dataset_manifest = scan_dataset_manifest(&service, dataset_name).await;

    let prop_col = dataset_manifest
        .column_by_name("property:text_log:TextLog:text")
        .expect("property column should exist")
        .downcast_array_ref::<ListArray>()
        .expect("property column should be a list array");

    let inner_array = prop_col.value(0);
    let string_array = inner_array
        .downcast_array_ref::<StringArray>()
        .expect("inner array should be string array");
    let text = string_array.value(0);

    assert_eq!(
        text, "second",
        "property should be 'second' after Overwrite behavior"
    );
}

/// Tests that intra-request duplicates always fail with `InvalidArgument`, regardless of
/// the `IfDuplicateBehavior` flag.
///
/// Intra-request duplicates are multiple data sources within a single registration request
/// that resolve to the same `(partition_id, layer)` pair. The `IfDuplicateBehavior` flag
/// only affects cross-request duplicates (conflicts with already-registered segments).
pub async fn register_intra_request_duplicates(service: impl RerunCloudService) {
    for on_duplicate in [
        IfDuplicateBehavior::Error,
        IfDuplicateBehavior::Skip,
        IfDuplicateBehavior::Overwrite,
    ] {
        let dataset_name = format!("intra_request_dup_{on_duplicate:?}_test");
        service.create_dataset_entry_with_name(&dataset_name).await;

        // Create two RRD files that will have the same partition ID
        let data_source_def = DataSourcesDefinition::new_with_tuid_prefix(
            1,
            [
                LayerDefinition::properties(
                    "segment1",
                    [prop(
                        "text_log",
                        re_sdk_types::archetypes::TextLog::new("first"),
                    )],
                ),
                LayerDefinition::properties(
                    "segment1", // DUPLICATE
                    [prop(
                        "text_log",
                        re_sdk_types::archetypes::TextLog::new("second"),
                    )],
                ),
            ],
        );

        let request = cloud_ext::RegisterWithDatasetRequest {
            data_sources: data_source_def.to_data_sources_ext(),
            on_duplicate,
        };

        let result = service
            .register_with_dataset(
                tonic::Request::new(request.into()).with_entry_name(entry_name(&dataset_name)),
            )
            .await;

        assert!(
            result.is_err(),
            "registration with intra-request duplicates should fail for {on_duplicate:?}"
        );

        let err = result.unwrap_err();
        assert_eq!(
            err.code(),
            tonic::Code::InvalidArgument,
            "intra-request duplicates should return InvalidArgument error for {on_duplicate:?}"
        );

        let msg = err.message().to_owned();
        assert!(
            msg.contains("duplicate segment layers in request"),
            "error message should mention duplicate segment layers for {on_duplicate:?}: {msg}"
        );
        assert!(
            msg.contains("segment id:"),
            "error message should contain segment id for {on_duplicate:?}: {msg}"
        );
        assert!(
            msg.contains("layer name:"),
            "error message should contain layer name for {on_duplicate:?}: {msg}"
        );

        // Both URIs should be listed in the error message
        for source in data_source_def.to_data_sources_ext() {
            assert!(
                msg.contains(source.storage_url.as_str()),
                "error message should contain URI {url} for {on_duplicate:?}: {msg}",
                url = source.storage_url,
            );
        }
    }
}

/// Tests that registering with an empty data sources list is rejected.
pub async fn register_empty_request(service: impl RerunCloudService) {
    let dataset_name = "empty_request_test";
    service.create_dataset_entry_with_name(dataset_name).await;

    // Register with empty data sources
    let request = cloud_ext::RegisterWithDatasetRequest {
        data_sources: vec![],
        on_duplicate: IfDuplicateBehavior::Error,
    };

    let result = service
        .register_with_dataset(
            tonic::Request::new(request.into()).with_entry_name(entry_name(dataset_name)),
        )
        .await;

    assert!(
        result.is_err(),
        "empty registration request should be rejected"
    );
    assert_eq!(
        result.unwrap_err().code(),
        tonic::Code::InvalidArgument,
        "empty registration should return InvalidArgument"
    );
}

/// Tests that a registration where all partitions are skipped (cross-request duplicates)
/// returns an empty response successfully.
pub async fn register_fully_skipped(service: impl RerunCloudService) {
    let dataset_name = "fully_skipped_test";
    service.create_dataset_entry_with_name(dataset_name).await;

    // First registration
    let data_sources_def = DataSourcesDefinition::new_with_tuid_prefix(
        1,
        [LayerDefinition::properties(
            "segment1",
            [prop(
                "text_log",
                re_sdk_types::archetypes::TextLog::new("first"),
            )],
        )],
    );

    service
        .register_with_dataset_name_blocking_with_behavior(
            dataset_name,
            data_sources_def.to_data_sources(),
            IfDuplicateBehavior::Skip,
        )
        .await;

    // Second registration with same partition - should be fully skipped
    let request = cloud_ext::RegisterWithDatasetRequest {
        data_sources: data_sources_def.to_data_sources_ext(),
        on_duplicate: IfDuplicateBehavior::Skip,
    };

    let result = service
        .register_with_dataset(
            tonic::Request::new(request.into()).with_entry_name(entry_name(dataset_name)),
        )
        .await;

    assert!(
        result.is_ok(),
        "fully skipped registration should succeed with empty response"
    );
}

// ---

async fn scan_dataset_manifest(
    service: &impl RerunCloudService,
    dataset_name: &str,
) -> RecordBatch {
    let responses: Vec<_> = service
        .scan_dataset_manifest(
            tonic::Request::new(ScanDatasetManifestRequest {
                columns: vec![], // all of them
            })
            .with_entry_name(entry_name(dataset_name)),
        )
        .await
        .unwrap()
        .into_inner()
        .try_collect()
        .await
        .unwrap();

    let batches: Vec<RecordBatch> = responses
        .into_iter()
        .map(|resp| resp.data.unwrap().try_into().unwrap())
        .collect_vec();

    arrow::compute::concat_batches(
        batches
            .first()
            .expect("there should be at least one batch")
            .schema_ref(),
        &batches,
    )
    .unwrap()
}

// ---

/// Test that two RRDs with conflicting data schemas cannot be registered together.
pub async fn register_conflicting_schema(service: impl RerunCloudService) {
    let results = register_and_wait_for_task_result(
        &service,
        "test_conflicting_schema",
        DataSourcesDefinition::new_with_tuid_prefix(
            1,
            [
                // Float64
                LayerDefinition::static_components(
                    "segment1",
                    [
                        (
                            EntityPath::from("/data"),
                            Box::new(AnyValues::default().with_component_from_data(
                                "test",
                                Arc::new(Float64Array::from(vec![1.0, 2.0, 3.0])),
                            )) as Box<dyn AsComponents>,
                        ), //
                        (
                            "/__properties/prop".into(),
                            Box::new(AnyValues::default().with_component_from_data(
                                "test",
                                Arc::new(Float64Array::from(vec![1.0, 2.0, 3.0])),
                            )) as Box<dyn AsComponents>,
                        ),
                    ],
                ),
                // Float32
                LayerDefinition::static_components(
                    "segment2",
                    [
                        (
                            EntityPath::from("/data"),
                            Box::new(AnyValues::default().with_component_from_data(
                                "test",
                                Arc::new(Float32Array::from(vec![1.0f32, 2.0, 3.0])),
                            )) as Box<dyn AsComponents>,
                        ), //
                        (
                            "/__properties/prop".into(),
                            Box::new(AnyValues::default().with_component_from_data(
                                "test",
                                Arc::new(Float64Array::from(vec![4.0, 5.0, 6.0])),
                            )) as Box<dyn AsComponents>,
                        ),
                    ],
                ),
            ],
        ),
    )
    .await;

    let failed_tasks: Vec<_> = results.iter().filter(|r| r.status != "success").collect();
    assert_eq!(
        failed_tasks.len(),
        1,
        "Expected exactly one task to fail with schema conflict"
    );

    let error_message = &failed_tasks[0].message;
    assert!(
        error_message
            .to_lowercase()
            .contains("schema incompatibility "),
        "error should mention schema conflict, got: {error_message}"
    );

    scan_segment_table_and_snapshot(
        &service,
        "test_conflicting_schema",
        "segment1_props_should_be_there",
    )
    .await;
}

/// Test that two RRDs with conflicting property schemas cannot be registered together.
pub async fn register_conflicting_property_schema(service: impl RerunCloudService) {
    let results = register_and_wait_for_task_result(
        &service,
        "test_conflicting_property_schema",
        DataSourcesDefinition::new_with_tuid_prefix(
            2,
            [
                // Float64
                LayerDefinition::properties(
                    "segment1",
                    [(
                        "prop".to_owned(),
                        Box::new(AnyValues::default().with_component_from_data(
                            "test",
                            Arc::new(Float64Array::from(vec![1.0, 2.0, 3.0])),
                        )) as Box<dyn AsComponents>,
                    )],
                ),
                // Float32
                LayerDefinition::properties(
                    "segment2",
                    [(
                        "prop".to_owned(),
                        Box::new(AnyValues::default().with_component_from_data(
                            "test",
                            Arc::new(Float32Array::from(vec![1.0f32, 2.0, 3.0])),
                        )) as Box<dyn AsComponents>,
                    )],
                ),
            ],
        ),
    )
    .await;

    let failed_tasks: Vec<_> = results.iter().filter(|r| r.status != "success").collect();
    assert_eq!(
        failed_tasks.len(),
        1,
        "Expected exactly one task to fail with schema conflict"
    );

    let error_message = &failed_tasks[0].message;
    assert!(
        error_message
            .to_lowercase()
            .contains("schema incompatibility "),
        "error should mention schema conflict, got: {error_message}"
    );
}

/// Test that segments with failed registrations do NOT appear in the segment table.
///
/// This test registers two segments with conflicting schemas in SEPARATE requests.
/// The first registration should succeed, and the second should fail with a schema conflict.
/// The segment table should only show the first (successful) segment.
pub async fn register_conflicting_schema_filters_segment_table(service: impl RerunCloudService) {
    use super::common::register_and_wait;

    let dataset_name = "test_conflicting_schema_filters_segment_table";
    service.create_dataset_entry_with_name(dataset_name).await;

    // First registration: segment1 with Float64 - should succeed
    let first_def = DataSourcesDefinition::new_with_tuid_prefix(
        100,
        [LayerDefinition::static_components(
            "segment1",
            [(
                EntityPath::from("/data"),
                Box::new(AnyValues::default().with_component_from_data(
                    "test",
                    Arc::new(Float64Array::from(vec![1.0, 2.0, 3.0])),
                )) as Box<dyn AsComponents>,
            )],
        )],
    );
    service
        .register_with_dataset_name_blocking(dataset_name, first_def.to_data_sources())
        .await;

    // Second registration: segment2 with Float32 - will fail due to schema conflict
    let second_def = DataSourcesDefinition::new_with_tuid_prefix(
        101,
        [LayerDefinition::static_components(
            "segment2",
            [(
                EntityPath::from("/data"),
                Box::new(AnyValues::default().with_component_from_data(
                    "test",
                    Arc::new(Float32Array::from(vec![1.0f32, 2.0, 3.0])),
                )) as Box<dyn AsComponents>,
            )],
        )],
    );
    let request = tonic::Request::new(re_protos::cloud::v1alpha1::RegisterWithDatasetRequest {
        data_sources: second_def.to_data_sources(),
        on_duplicate: IfDuplicateBehavior::Error as i32,
    })
    .with_entry_name(entry_name(dataset_name));

    // This registration should fail due to schema conflict
    let task_results = register_and_wait(&service, request).await;
    assert_task_failed(&task_results, "schema");

    // Verify the segment table only contains segment1 (the successful one)
    let segment_table = scan_segment_table(&service, dataset_name).await;

    let segment_id_col = ScanSegmentTableDataframe::COLUMN_RERUN_SEGMENT_ID
        .extract(&segment_table)
        .expect("segment_id column expected");

    let segment_ids: Vec<&str> = segment_id_col.iter().collect();

    assert_eq!(
        segment_ids.len(),
        1,
        "Segment table should only contain 1 segment (segment1), got: {segment_ids:?}"
    );
    assert_eq!(
        segment_ids[0], "segment1",
        "Segment table should contain 'segment1', got: {segment_ids:?}"
    );
}

/// Test that layers with failed registrations do NOT appear in the segment table.
///
/// This test registers two layers for the same segment with conflicting schemas in SEPARATE requests.
/// The first registration should succeed, and the second should fail with a schema conflict.
/// The segment table should show the segment with only the first (successful) layer.
pub async fn register_conflicting_schema_same_segment_filters_layer(
    service: impl RerunCloudService,
) {
    use super::common::register_and_wait;

    let dataset_name = "test_conflicting_schema_same_segment_filters_layer";
    service.create_dataset_entry_with_name(dataset_name).await;

    // First registration: segment1/base with Float64 - should succeed
    let first_def = DataSourcesDefinition::new_with_tuid_prefix(
        200,
        [LayerDefinition::static_components(
            "segment1",
            [(
                EntityPath::from("/data"),
                Box::new(AnyValues::default().with_component_from_data(
                    "test",
                    Arc::new(Float64Array::from(vec![1.0, 2.0, 3.0])),
                )) as Box<dyn AsComponents>,
            )],
        )
        .layer_name("base")],
    );
    service
        .register_with_dataset_name_blocking(dataset_name, first_def.to_data_sources())
        .await;

    // Second registration: segment1/extra with Float32 - will fail due to schema conflict
    let second_def = DataSourcesDefinition::new_with_tuid_prefix(
        201,
        [LayerDefinition::static_components(
            "segment1",
            [(
                EntityPath::from("/data"),
                Box::new(AnyValues::default().with_component_from_data(
                    "test",
                    Arc::new(Float32Array::from(vec![1.0f32, 2.0, 3.0])),
                )) as Box<dyn AsComponents>,
            )],
        )
        .layer_name("extra")],
    );
    let request = tonic::Request::new(re_protos::cloud::v1alpha1::RegisterWithDatasetRequest {
        data_sources: second_def.to_data_sources(),
        on_duplicate: IfDuplicateBehavior::Error as i32,
    })
    .with_entry_name(entry_name(dataset_name));

    // This registration should fail due to schema conflict
    let task_results = register_and_wait(&service, request).await;
    assert_task_failed(&task_results, "schema");

    // Verify the segment table shows segment1 with only the base layer
    let segment_table = scan_segment_table(&service, dataset_name).await;

    let segment_id_col = ScanSegmentTableDataframe::COLUMN_RERUN_SEGMENT_ID
        .extract(&segment_table)
        .expect("segment_id column expected");

    let layer_names_col = ScanSegmentTableDataframe::COLUMN_RERUN_LAYER_NAMES
        .extract(&segment_table)
        .expect("layer_names column expected");

    assert_eq!(
        segment_id_col.len(),
        1,
        "Segment table should only contain 1 segment"
    );
    assert_eq!(
        &segment_id_col[0], "segment1",
        "Segment table should contain 'segment1'"
    );

    let layers: Vec<&str> = layer_names_col.value(0).collect();

    assert_eq!(
        layers,
        vec!["base"],
        "Segment should only have 'base' layer (the successful one), got: {layers:?}"
    );
}

/// Helper to assert that at least one task failed with a message containing the expected substring.
fn assert_task_failed(task_results: &[RecordBatch], expected_message_substring: &str) {
    let mut found_failure = false;
    let mut failure_message = String::new();

    for batch in task_results {
        let statuses = cloud_ext::QueryTasksDataframe::COLUMN_EXEC_STATUS
            .extract(batch)
            .expect("valid exec_status column");
        let msgs = cloud_ext::QueryTasksDataframe::COLUMN_MSGS
            .extract(batch)
            .expect("valid msgs column");

        for (status, msg) in izip!(&statuses, &msgs) {
            if status != "success" {
                found_failure = true;
                failure_message = msg.unwrap_or_default().to_owned();
                break;
            }
        }
        if found_failure {
            break;
        }
    }

    assert!(
        found_failure,
        "Expected at least one task to fail, but all tasks succeeded"
    );
    assert!(
        failure_message
            .to_lowercase()
            .contains(&expected_message_substring.to_lowercase()),
        "Expected failure message to contain '{expected_message_substring}', got: {failure_message}"
    );
}

/// Helper to scan the segment table for a dataset.
async fn scan_segment_table(service: &impl RerunCloudService, dataset_name: &str) -> RecordBatch {
    let responses: Vec<_> = service
        .scan_segment_table(
            tonic::Request::new(ScanSegmentTableRequest {
                columns: vec![], // all of them
            })
            .with_entry_name(entry_name(dataset_name)),
        )
        .await
        .unwrap()
        .into_inner()
        .try_collect()
        .await
        .unwrap();

    let batches: Vec<RecordBatch> = responses
        .into_iter()
        .map(|resp| resp.data.unwrap().try_into().unwrap())
        .collect_vec();

    // Handle empty responses by returning an empty batch with the expected schema
    if batches.is_empty() {
        return RecordBatch::new_empty(Arc::new(ScanSegmentTableDataframe::min_schema()));
    }

    arrow::compute::concat_batches(batches.first().unwrap().schema_ref(), &batches).unwrap()
}

// ---

/// Result of a registered task, including which segments/layers were registered.
#[derive(Debug)]
struct TaskResult {
    #[expect(dead_code)]
    task_id: TaskId,
    status: String,
    message: String,
    #[expect(dead_code)]
    layers: Vec<(SegmentId, LayerName)>,
}

/// Helper to register data sources and wait for task completion.
/// Returns a vec of task results, one per unique task.
async fn register_and_wait_for_task_result(
    service: &impl RerunCloudService,
    dataset_name: &str,
    data_sources_def: DataSourcesDefinition,
) -> Vec<TaskResult> {
    use futures::StreamExt as _;
    use re_protos::cloud::v1alpha1::QueryTasksOnCompletionRequest;
    use re_protos::common::v1alpha1::TaskId;
    use std::collections::HashMap;

    service.create_dataset_entry_with_name(dataset_name).await;

    let request = tonic::Request::new(re_protos::cloud::v1alpha1::RegisterWithDatasetRequest {
        data_sources: data_sources_def.to_data_sources(),
        on_duplicate: IfDuplicateBehavior::Error as i32,
    })
    .with_entry_name(entry_name(dataset_name));

    let resp = service
        .register_with_dataset(request)
        .await
        .expect("registration should succeed");

    let batch: RecordBatch = resp
        .into_inner()
        .data
        .expect("data expected")
        .try_into()
        .expect("record batch expected");

    // Extract task IDs and group segments by task
    let ext::RegisterWithDatasetDataframe {
        rerun_segment_id,
        rerun_segment_layer,
        rerun_task_id,
        ..
    } = ext::RegisterWithDatasetDataframe::try_from(batch)
        .expect("valid RegisterWithDataset response dataframe");

    // Group (segment_id, layer) by task_id
    let mut task_layers: HashMap<TaskId, Vec<(SegmentId, LayerName)>> = HashMap::default();
    for (task_id, segment_id, layer_name) in izip!(
        rerun_task_id.into_iter_owned(),
        rerun_segment_id.into_iter_owned(),
        rerun_segment_layer.into_iter_owned()
    ) {
        task_layers
            .entry(task_id)
            .or_default()
            .push((segment_id, layer_name));
    }

    let task_ids: Vec<TaskId> = task_layers.keys().cloned().collect();

    // Wait for task completion
    let query_results: Vec<RecordBatch> = service
        .query_tasks_on_completion(tonic::Request::new(QueryTasksOnCompletionRequest {
            ids: task_ids,
            timeout: Some(prost_types::Duration {
                seconds: 20,
                nanos: 0,
            }),
        }))
        .await
        .expect("should get query results")
        .into_inner()
        .collect::<Vec<_>>()
        .await
        .into_iter()
        .map(|resp| {
            resp.expect("Failed to get task completion response")
                .data
                .expect("Expected response data")
                .try_into()
                .expect("Failed to decode response data")
        })
        .collect();

    // Build TaskResult for each task
    let mut results = Vec::new();
    for batch in &query_results {
        let cloud_ext::QueryTasksDataframe {
            task_id,
            exec_status,
            msgs,
            ..
        } = cloud_ext::QueryTasksDataframe::try_from(batch).expect("valid QueryTasks dataframe");

        for (task_id, status, message) in izip!(
            task_id.into_iter_owned(),
            exec_status.into_iter_owned(),
            msgs.into_iter_owned()
        ) {
            let message = message.unwrap_or_default();
            let layers = task_layers.remove(&task_id).unwrap_or_default();

            results.push(TaskResult {
                task_id,
                status,
                message,
                layers,
            });
        }
    }

    results
}

// ---

async fn scan_segment_table_and_snapshot(
    service: &impl RerunCloudService,
    dataset_name: &str,
    snapshot_name: &str,
) -> RecordBatch {
    let responses: Vec<_> = service
        .scan_segment_table(
            tonic::Request::new(ScanSegmentTableRequest {
                columns: vec![], // all of them
            })
            .with_entry_name(entry_name(dataset_name)),
        )
        .await
        .unwrap()
        .into_inner()
        .try_collect()
        .await
        .unwrap();

    let batches: Vec<RecordBatch> = responses
        .into_iter()
        .map(|resp| resp.data.unwrap().try_into().unwrap())
        .collect_vec();

    let batch = arrow::compute::concat_batches(
        batches
            .first()
            .expect("there should be at least one batch")
            .schema_ref(),
        &batches,
    )
    .unwrap();

    // check that the _advertised_ schema is consistent with the actual data.
    let alleged_schema: Schema = service
        .get_segment_table_schema(
            tonic::Request::new(GetSegmentTableSchemaRequest {})
                .with_entry_name(entry_name(dataset_name)),
        )
        .await
        .unwrap()
        .into_inner()
        .schema
        .unwrap()
        .try_into()
        .unwrap();

    // Note: we check fields only because some schema-level sorbet metadata is injected in one of
    // the paths and not the other. Anyway, that's what matters.
    assert_eq!(
        alleged_schema.fields(),
        batch.schema_ref().fields(),
        "The actual schema is not consistent with the schema advertised by \
        `get_segment_table_schema`.\n\nActual:\n{}\n\nAlleged:\n{}\n",
        batch.schema().format_snapshot(),
        alleged_schema.format_snapshot(),
    );

    let required_fields = ScanSegmentTableDataframe::min_schema().fields().to_vec();
    assert!(
        batch.schema().fields().contains_unordered(&required_fields),
        "the schema should contain all the required fields, but it doesn't",
    );

    let unstable_column_names = vec![
        ScanSegmentTableDataframe::COLUMN_RERUN_STORAGE_URLS_NAME,
        ScanSegmentTableDataframe::COLUMN_RERUN_SIZE_BYTES_NAME,
        ScanSegmentTableDataframe::COLUMN_RERUN_LAST_UPDATED_AT_NAME,
    ];
    let filtered_batch = batch
        .remove_columns(&unstable_column_names)
        .auto_sort_rows()
        .unwrap()
        .sort_property_columns()
        .sort_index_columns();

    insta::assert_snapshot!(
        format!("{snapshot_name}_segments_schema"),
        batch.format_schema_snapshot()
    );
    insta::assert_snapshot!(
        format!("{snapshot_name}_segments_data"),
        filtered_batch.format_snapshot(false)
    );

    batch
}

async fn scan_dataset_manifest_and_snapshot(
    service: &impl RerunCloudService,
    dataset_name: &str,
    snapshot_name: &str,
) -> RecordBatch {
    let responses: Vec<_> = service
        .scan_dataset_manifest(
            tonic::Request::new(ScanDatasetManifestRequest {
                columns: vec![], // all of them
            })
            .with_entry_name(entry_name(dataset_name)),
        )
        .await
        .unwrap()
        .into_inner()
        .try_collect()
        .await
        .unwrap();

    let batches: Vec<RecordBatch> = responses
        .into_iter()
        .map(|resp| resp.data.unwrap().try_into().unwrap())
        .collect_vec();

    let batch = arrow::compute::concat_batches(
        batches
            .first()
            .expect("there should be at least one batch")
            .schema_ref(),
        &batches,
    )
    .unwrap();

    // check that the _advertised_ schema is consistent with the actual data.
    let alleged_schema: Schema = service
        .get_dataset_manifest_schema(
            tonic::Request::new(GetDatasetManifestSchemaRequest {})
                .with_entry_name(entry_name(dataset_name)),
        )
        .await
        .unwrap()
        .into_inner()
        .schema
        .unwrap()
        .try_into()
        .unwrap();

    // Note: we check fields only because some schema-level sorbet metadata is injected in one of
    // the paths and not the other. Anyway, that's what matters.
    assert_eq!(
        alleged_schema.fields(),
        batch.schema_ref().fields(),
        "The actual schema is not consistent with the schema advertised by \
        `get_dataset_manifest_schema`.\n\nActual:\n{}\n\nAlleged:\n{}\n",
        batch.schema().format_snapshot(),
        alleged_schema.format_snapshot(),
    );

    let required_fields = ScanDatasetManifestDataframe::min_schema().fields().to_vec();
    assert!(
        batch.schema().fields().contains_unordered(&required_fields),
        "the schema should contain all the required fields, but it doesn't",
    );

    let unstable_column_names = vec![
        // implementation-dependent
        ScanDatasetManifestDataframe::COLUMN_RERUN_STORAGE_URL_NAME,
        ScanDatasetManifestDataframe::COLUMN_RERUN_SIZE_BYTES_NAME,
        // unstable
        ScanDatasetManifestDataframe::COLUMN_RERUN_LAST_UPDATED_AT_NAME,
        ScanDatasetManifestDataframe::COLUMN_RERUN_REGISTRATION_TIME_NAME,
    ];
    let filtered_batch = batch
        .remove_columns(&unstable_column_names)
        .auto_sort_rows()
        .unwrap()
        .sort_property_columns();

    insta::assert_snapshot!(
        format!("{snapshot_name}_manifest_schema"),
        batch.format_schema_snapshot()
    );
    insta::assert_snapshot!(
        format!("{snapshot_name}_manifest_data"),
        filtered_batch.format_snapshot(false)
    );

    batch
}