示例#1
0
    def modify_uuid(self, target_uuid: str):
        """
        Modify the dataset uuid and depending metadata:
        - paths to partitioning files
        - path to index files
        Parameters
        ----------
        target_uuid: str
            Modified dataset UUID.
        Returns
        -------
        DatasetMetadataBuilder
            modified builder object
        """

        # modify file names in partition metadata
        modified_partitions = {}
        for p_key, p in self.partitions.items():
            pdict = p.to_dict()
            for table_key, table_file in pdict["files"].items():
                if table_file.startswith(f"{self.uuid}/"):
                    pdict["files"][table_key] = table_file.replace(
                        self.uuid, target_uuid, 1)
            modified_partitions[p_key] = Partition.from_dict(p_key, pdict)

        self.partitions = modified_partitions

        for i_key, i in self.indices.items():
            if (isinstance(i, ExplicitSecondaryIndex)
                    and i.index_storage_key is not None):
                i.index_storage_key = i.index_storage_key.replace(
                    self.uuid, target_uuid, 1)
        self.uuid = target_uuid
        return self
示例#2
0
    def from_dict(dct: Dict, explicit_partitions: bool = True):
        """
        Load dataset metadata from a dictionary.

        This must have no external references. Otherwise use ``load_from_dict``
        to have them resolved automatically.
        """

        # Use the builder class for reconstruction to have a single point for metadata version changes
        builder = DatasetMetadataBuilder(
            uuid=dct[naming.UUID_KEY],
            metadata_version=dct[naming.METADATA_VERSION_KEY],
            explicit_partitions=explicit_partitions,
            partition_keys=dct.get("partition_keys", None),
            table_meta=dct.get("table_meta", None),
        )

        for key, value in dct.get("metadata", {}).items():
            builder.add_metadata(key, value)
        for partition_label, part_dct in dct.get("partitions", {}).items():
            builder.add_partition(
                partition_label, Partition.from_dict(partition_label,
                                                     part_dct))
        for column, index_dct in dct.get("indices", {}).items():
            if isinstance(index_dct, IndexBase):
                builder.add_embedded_index(column, index_dct)
            else:
                builder.add_embedded_index(
                    column, ExplicitSecondaryIndex.from_v2(column, index_dct))
        return builder.to_dataset()
示例#3
0
def test_raise_on_erroneous_input():
    with pytest.raises(ValueError):
        Partition.from_dict(label="label",
                            dct="some_not_supported_external_ref")
示例#4
0
def test_roundtrip():
    expected = {"files": {"Queejeb3": "file.parquet"}}
    result = Partition.from_dict("partition_label", expected).to_dict()
    assert expected == result
示例#5
0
def test_roundtrip_empty_metadata():
    _input = {"files": {"Queejeb3": "file.parquet"}}
    expected = {"files": {"Queejeb3": "file.parquet"}}
    result = Partition.from_dict("partition_label", _input).to_dict()
    assert expected == result