Skip to content

climate_ref.doctor.context #

The deployment a check runs against.

EMPTY_CATALOG = pd.DataFrame() module-attribute #

Stands in for a source type with nothing ingested.

DoctorContext #

The deployment being checked.

Providers and catalogs are loaded lazily so a check that does not need them does not pay for them, and so a failure to load one provider does not stop the other checks.

Source code in packages/climate-ref/src/climate_ref/doctor/context.py 
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
@define
class DoctorContext:
    """
    The deployment being checked.

    Providers and catalogs are loaded lazily so a check that does not need them does not pay for them,
    and so a failure to load one provider does not stop the other checks.
    """

    config: Config | None
    database: Database | None
    _providers: list[DiagnosticProvider] | None = field(default=None, alias="_providers")
    _catalogs: dict[SourceDatasetType, pd.DataFrame] = field(factory=dict, alias="_catalogs")

    @classmethod
    def from_catalogs(
        cls,
        catalogs: dict[SourceDatasetType, pd.DataFrame],
        providers: Iterable[DiagnosticProvider],
    ) -> "DoctorContext":
        """
        Build a context from catalogs already in hand, with no database behind it.

        Source types absent from ``catalogs`` are treated as having nothing ingested,
        so every check can run without reaching for a database that is not there.

        Parameters
        ----------
        catalogs
            The catalogs to check, keyed by source type.
        providers
            The providers to treat as enabled.

        Returns
        -------
        :
            A context backed by nothing but the supplied catalogs and providers.
        """
        complete = {
            source_type: catalogs.get(source_type, EMPTY_CATALOG) for source_type in SourceDatasetType
        }
        return cls(config=None, database=None, _providers=list(providers), _catalogs=complete)

    @property
    def providers(self) -> list[DiagnosticProvider]:
        """The diagnostic providers this deployment has enabled."""
        if self._providers is None:
            from climate_ref.provider_registry import ProviderRegistry  # noqa: PLC0415

            if self.config is None or self.database is None:
                raise ValueError("This context has no configuration to load providers from")
            registry = ProviderRegistry.build_from_config(
                self.config, self.database, configure=False, register=False
            )
            self._providers = list(registry.providers)
        return self._providers

    def data_catalog(self, source_type: SourceDatasetType) -> DataCatalog:
        """
        Load the ingested catalog as the solver sees it.

        The adapter and database are carried through where there are any,
        so an unfinalised dataset can still be finalised during a solve.

        Parameters
        ----------
        source_type
            The source type to load.

        Returns
        -------
        :
            A catalog the solver can be run against.
        """
        frame = self.catalog(source_type)
        if self.database is None:
            return DataCatalog.from_frame(frame)
        return DataCatalog(database=self.database, adapter=get_dataset_adapter(source_type.value), df=frame)

    def catalog(self, source_type: SourceDatasetType) -> pd.DataFrame:
        """
        Load the ingested catalog for a source type, one row per file.

        Parameters
        ----------
        source_type
            The source type to load.

        Returns
        -------
        :
            The catalog, or an empty frame when nothing of that type has been ingested.
        """
        if source_type not in self._catalogs:
            if self.database is None:
                raise ValueError("This context has no database to load a catalog from")
            adapter = get_dataset_adapter(source_type.value)
            self._catalogs[source_type] = adapter.load_catalog(self.database)
        return self._catalogs[source_type]

providers property #

The diagnostic providers this deployment has enabled.

catalog(source_type) #

Load the ingested catalog for a source type, one row per file.

Parameters:

Name Type Description Default
source_type SourceDatasetType

The source type to load.

 required

Returns:

Type Description
DataFrame

The catalog, or an empty frame when nothing of that type has been ingested.
Source code in packages/climate-ref/src/climate_ref/doctor/context.py 
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
def catalog(self, source_type: SourceDatasetType) -> pd.DataFrame:
    """
    Load the ingested catalog for a source type, one row per file.

    Parameters
    ----------
    source_type
        The source type to load.

    Returns
    -------
    :
        The catalog, or an empty frame when nothing of that type has been ingested.
    """
    if source_type not in self._catalogs:
        if self.database is None:
            raise ValueError("This context has no database to load a catalog from")
        adapter = get_dataset_adapter(source_type.value)
        self._catalogs[source_type] = adapter.load_catalog(self.database)
    return self._catalogs[source_type]

data_catalog(source_type) #

Load the ingested catalog as the solver sees it.

The adapter and database are carried through where there are any, so an unfinalised dataset can still be finalised during a solve.

Parameters:

Name Type Description Default
source_type SourceDatasetType

The source type to load.

 required

Returns:

Type Description
DataCatalog

A catalog the solver can be run against.
Source code in packages/climate-ref/src/climate_ref/doctor/context.py 
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
def data_catalog(self, source_type: SourceDatasetType) -> DataCatalog:
    """
    Load the ingested catalog as the solver sees it.

    The adapter and database are carried through where there are any,
    so an unfinalised dataset can still be finalised during a solve.

    Parameters
    ----------
    source_type
        The source type to load.

    Returns
    -------
    :
        A catalog the solver can be run against.
    """
    frame = self.catalog(source_type)
    if self.database is None:
        return DataCatalog.from_frame(frame)
    return DataCatalog(database=self.database, adapter=get_dataset_adapter(source_type.value), df=frame)

from_catalogs(catalogs, providers) classmethod #

Build a context from catalogs already in hand, with no database behind it.

Source types absent from catalogs are treated as having nothing ingested, so every check can run without reaching for a database that is not there.

Parameters:

Name Type Description Default
catalogs dict[SourceDatasetType, DataFrame]

The catalogs to check, keyed by source type.

 required
providers Iterable[DiagnosticProvider]

The providers to treat as enabled.

 required

Returns:

Type Description
DoctorContext

A context backed by nothing but the supplied catalogs and providers.
Source code in packages/climate-ref/src/climate_ref/doctor/context.py 
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
@classmethod
def from_catalogs(
    cls,
    catalogs: dict[SourceDatasetType, pd.DataFrame],
    providers: Iterable[DiagnosticProvider],
) -> "DoctorContext":
    """
    Build a context from catalogs already in hand, with no database behind it.

    Source types absent from ``catalogs`` are treated as having nothing ingested,
    so every check can run without reaching for a database that is not there.

    Parameters
    ----------
    catalogs
        The catalogs to check, keyed by source type.
    providers
        The providers to treat as enabled.

    Returns
    -------
    :
        A context backed by nothing but the supplied catalogs and providers.
    """
    complete = {
        source_type: catalogs.get(source_type, EMPTY_CATALOG) for source_type in SourceDatasetType
    }
    return cls(config=None, database=None, _providers=list(providers), _catalogs=complete)