From 6cc1735fa692f02e880af45081c6cdef1e2fe154 Mon Sep 17 00:00:00 2001 From: Fokko Date: Tue, 13 May 2025 18:21:28 +0200 Subject: [PATCH 1/3] Add `snapshot-loading-mode` to RESTCatalog This will allow to set the snapshots to be send back. In case of refs, only the snapshots referenced by branches or tags will be returned: https://github.com/apache/iceberg/blob/5d2230ead79da64a8c871a02eb1304a94aaece5c/open-api/rest-catalog-open-api.yaml#L954-L956 --- pyiceberg/catalog/rest/__init__.py | 14 ++++++++-- tests/catalog/test_rest.py | 43 +++++++++++++++++++++++++++++- 2 files changed, 54 insertions(+), 3 deletions(-) diff --git a/pyiceberg/catalog/rest/__init__.py b/pyiceberg/catalog/rest/__init__.py index 8ee9e5fdc9..633ef6498b 100644 --- a/pyiceberg/catalog/rest/__init__.py +++ b/pyiceberg/catalog/rest/__init__.py @@ -138,6 +138,7 @@ class IdentifierKind(Enum): SIGV4_REGION = "rest.signing-region" SIGV4_SERVICE = "rest.signing-name" OAUTH2_SERVER_URI = "oauth2-server-uri" +SNAPSHOT_LOADING_MODE = "snapshot-loading-mode" NAMESPACE_SEPARATOR = b"\x1f".decode(UTF8) @@ -678,7 +679,16 @@ def list_tables(self, namespace: Union[str, Identifier]) -> List[Identifier]: @retry(**_RETRY_ARGS) def load_table(self, identifier: Union[str, Identifier]) -> Table: - response = self._session.get(self.url(Endpoints.load_table, prefixed=True, **self._split_identifier_for_path(identifier))) + params = {} + if mode := self.properties.get(SNAPSHOT_LOADING_MODE): + if mode in {"all", "refs"}: + params["snapshots"] = mode + else: + raise ValueError("Invalid snapshot-loading-mode: {}") + + response = self._session.get( + self.url(Endpoints.load_table, prefixed=True, **self._split_identifier_for_path(identifier)), params=params + ) try: response.raise_for_status() except HTTPError as exc: @@ -816,7 +826,7 @@ def list_namespaces(self, namespace: Union[str, Identifier] = ()) -> List[Identi try: response.raise_for_status() except HTTPError as exc: - self._handle_non_200_response(exc, {}) + self._handle_non_200_response(exc, {404: NoSuchNamespaceError}) return ListNamespaceResponse.model_validate_json(response.text).namespaces diff --git a/tests/catalog/test_rest.py b/tests/catalog/test_rest.py index f2fc6ceb6b..dfc6a0e2fb 100644 --- a/tests/catalog/test_rest.py +++ b/tests/catalog/test_rest.py @@ -24,7 +24,7 @@ import pyiceberg from pyiceberg.catalog import PropertiesUpdateSummary, load_catalog -from pyiceberg.catalog.rest import OAUTH2_SERVER_URI, RestCatalog +from pyiceberg.catalog.rest import OAUTH2_SERVER_URI, SNAPSHOT_LOADING_MODE, RestCatalog from pyiceberg.exceptions import ( AuthorizationExpiredError, NamespaceAlreadyExistsError, @@ -555,6 +555,24 @@ def test_list_namespace_with_parent_200(rest_mock: Mocker) -> None: ] +def test_list_namespace_with_parent_404(rest_mock: Mocker) -> None: + rest_mock.get( + f"{TEST_URI}v1/namespaces?parent=some_namespace", + json={ + "error": { + "message": "Namespace provided in the `parent` query parameter is not found", + "type": "NoSuchNamespaceException", + "code": 404, + } + }, + status_code=404, + request_headers=TEST_HEADERS, + ) + + with pytest.raises(NoSuchNamespaceError): + RestCatalog("rest", uri=TEST_URI, token=TEST_TOKEN).list_namespaces(("some_namespace",)) + + @pytest.mark.filterwarnings( "ignore:Deprecated in 0.8.0, will be removed in 1.0.0. Iceberg REST client is missing the OAuth2 server URI:DeprecationWarning" ) @@ -835,6 +853,29 @@ def test_load_table_200(rest_mock: Mocker, example_table_metadata_with_snapshot_ assert actual == expected +def test_load_table_200_loading_mode( + rest_mock: Mocker, example_table_metadata_with_snapshot_v1_rest_json: Dict[str, Any] +) -> None: + rest_mock.get( + f"{TEST_URI}v1/namespaces/fokko/tables/table?snapshots=refs", + json=example_table_metadata_with_snapshot_v1_rest_json, + status_code=200, + request_headers=TEST_HEADERS, + ) + catalog = RestCatalog("rest", uri=TEST_URI, token=TEST_TOKEN, **{SNAPSHOT_LOADING_MODE: "refs"}) + actual = catalog.load_table(("fokko", "table")) + expected = Table( + identifier=("fokko", "table"), + metadata_location=example_table_metadata_with_snapshot_v1_rest_json["metadata-location"], + metadata=TableMetadataV1(**example_table_metadata_with_snapshot_v1_rest_json["metadata"]), + io=load_file_io(), + catalog=catalog, + ) + # First compare the dicts + assert actual.metadata.model_dump() == expected.metadata.model_dump() + assert actual == expected + + def test_load_table_honor_access_delegation( rest_mock: Mocker, example_table_metadata_with_snapshot_v1_rest_json: Dict[str, Any] ) -> None: From e3c1c1b1497f9cd27be2d4e54d1e5aec1e69c198 Mon Sep 17 00:00:00 2001 From: Fokko Date: Tue, 13 May 2025 18:26:57 +0200 Subject: [PATCH 2/3] Add docs --- mkdocs/docs/configuration.md | 1 + 1 file changed, 1 insertion(+) diff --git a/mkdocs/docs/configuration.md b/mkdocs/docs/configuration.md index 1e364a11fe..dc51f81434 100644 --- a/mkdocs/docs/configuration.md +++ b/mkdocs/docs/configuration.md @@ -332,6 +332,7 @@ catalog: | rest.signing-region | us-east-1 | The region to use when SigV4 signing a request | | rest.signing-name | execute-api | The service signing name to use when SigV4 signing a request | | oauth2-server-uri | | Authentication URL to use for client credentials authentication (default: uri + 'v1/oauth/tokens') | +| snapshot-loading-mode | refs | The snapshots to return in the body of the metadata. Setting the value to `all` would return the full set of snapshots currently valid for the table. Setting the value to `refs` would load all snapshots referenced by branches or tags. | From e22b7319d6ece5014bb9edcd5659a4bc1305e75d Mon Sep 17 00:00:00 2001 From: "dependabot[bot]" <49699333+dependabot[bot]@users.noreply.github.com> Date: Mon, 26 May 2025 01:36:13 +0000 Subject: [PATCH 3/3] Build: Bump mypy-boto3-glue from 1.38.12 to 1.38.22 Bumps [mypy-boto3-glue](https://github.com/youtype/mypy_boto3_builder) from 1.38.12 to 1.38.22. - [Release notes](https://github.com/youtype/mypy_boto3_builder/releases) - [Commits](https://github.com/youtype/mypy_boto3_builder/commits) --- updated-dependencies: - dependency-name: mypy-boto3-glue dependency-version: 1.38.22 dependency-type: direct:production update-type: version-update:semver-patch ... Signed-off-by: dependabot[bot] --- poetry.lock | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/poetry.lock b/poetry.lock index 849485ffbb..51293e0a1d 100644 --- a/poetry.lock +++ b/poetry.lock @@ -3076,15 +3076,15 @@ typing-extensions = {version = ">=4.1.0", markers = "python_version < \"3.11\""} [[package]] name = "mypy-boto3-glue" -version = "1.38.12" -description = "Type annotations for boto3 Glue 1.38.12 service generated with mypy-boto3-builder 8.11.0" +version = "1.38.22" +description = "Type annotations for boto3 Glue 1.38.22 service generated with mypy-boto3-builder 8.11.0" optional = true python-versions = ">=3.8" groups = ["main"] markers = "extra == \"glue\"" files = [ - {file = "mypy_boto3_glue-1.38.12-py3-none-any.whl", hash = "sha256:2fd93c0c941f5a603e6fba7c60d277053bf68a72df8e6e6532addef61267f892"}, - {file = "mypy_boto3_glue-1.38.12.tar.gz", hash = "sha256:d0c709de2f23b6a9277ec03bb0df4fefe20e32a6c748531049c17d8349bc5a09"}, + {file = "mypy_boto3_glue-1.38.22-py3-none-any.whl", hash = "sha256:4fe34c858cbee41e8ad30382305c01b0dd9c1da4c84f894860b9249ddabb4a58"}, + {file = "mypy_boto3_glue-1.38.22.tar.gz", hash = "sha256:a9c529fafaaa9845d39c3204b3fb6cbbb633fa747faf6a084a2b2a381ef12a2b"}, ] [package.dependencies]