pyjelly 0.7.1__cp311-cp311-macosx_11_0_x86_64.whl

pyjelly/jelly/rdf_pb2.pyi

@@ -0,0 +1,231 @@
+from google.protobuf.internal import containers as _containers
+from google.protobuf.internal import enum_type_wrapper as _enum_type_wrapper
+from google.protobuf import descriptor as _descriptor
+from google.protobuf import message as _message
+from collections.abc import Iterable as _Iterable, Mapping as _Mapping
+from typing import ClassVar as _ClassVar, Optional as _Optional, Union as _Union
+
+DESCRIPTOR: _descriptor.FileDescriptor
+
+class PhysicalStreamType(int, metaclass=_enum_type_wrapper.EnumTypeWrapper):
+    __slots__ = ()
+    PHYSICAL_STREAM_TYPE_UNSPECIFIED: _ClassVar[PhysicalStreamType]
+    PHYSICAL_STREAM_TYPE_TRIPLES: _ClassVar[PhysicalStreamType]
+    PHYSICAL_STREAM_TYPE_QUADS: _ClassVar[PhysicalStreamType]
+    PHYSICAL_STREAM_TYPE_GRAPHS: _ClassVar[PhysicalStreamType]
+
+class LogicalStreamType(int, metaclass=_enum_type_wrapper.EnumTypeWrapper):
+    __slots__ = ()
+    LOGICAL_STREAM_TYPE_UNSPECIFIED: _ClassVar[LogicalStreamType]
+    LOGICAL_STREAM_TYPE_FLAT_TRIPLES: _ClassVar[LogicalStreamType]
+    LOGICAL_STREAM_TYPE_FLAT_QUADS: _ClassVar[LogicalStreamType]
+    LOGICAL_STREAM_TYPE_GRAPHS: _ClassVar[LogicalStreamType]
+    LOGICAL_STREAM_TYPE_DATASETS: _ClassVar[LogicalStreamType]
+    LOGICAL_STREAM_TYPE_SUBJECT_GRAPHS: _ClassVar[LogicalStreamType]
+    LOGICAL_STREAM_TYPE_NAMED_GRAPHS: _ClassVar[LogicalStreamType]
+    LOGICAL_STREAM_TYPE_TIMESTAMPED_NAMED_GRAPHS: _ClassVar[LogicalStreamType]
+PHYSICAL_STREAM_TYPE_UNSPECIFIED: PhysicalStreamType
+PHYSICAL_STREAM_TYPE_TRIPLES: PhysicalStreamType
+PHYSICAL_STREAM_TYPE_QUADS: PhysicalStreamType
+PHYSICAL_STREAM_TYPE_GRAPHS: PhysicalStreamType
+LOGICAL_STREAM_TYPE_UNSPECIFIED: LogicalStreamType
+LOGICAL_STREAM_TYPE_FLAT_TRIPLES: LogicalStreamType
+LOGICAL_STREAM_TYPE_FLAT_QUADS: LogicalStreamType
+LOGICAL_STREAM_TYPE_GRAPHS: LogicalStreamType
+LOGICAL_STREAM_TYPE_DATASETS: LogicalStreamType
+LOGICAL_STREAM_TYPE_SUBJECT_GRAPHS: LogicalStreamType
+LOGICAL_STREAM_TYPE_NAMED_GRAPHS: LogicalStreamType
+LOGICAL_STREAM_TYPE_TIMESTAMPED_NAMED_GRAPHS: LogicalStreamType
+
+class RdfIri(_message.Message):
+    __slots__ = ("prefix_id", "name_id")
+    PREFIX_ID_FIELD_NUMBER: _ClassVar[int]
+    NAME_ID_FIELD_NUMBER: _ClassVar[int]
+    prefix_id: int
+    name_id: int
+    def __init__(self, prefix_id: _Optional[int] = ..., name_id: _Optional[int] = ...) -> None: ...
+
+class RdfLiteral(_message.Message):
+    __slots__ = ("lex", "langtag", "datatype")
+    LEX_FIELD_NUMBER: _ClassVar[int]
+    LANGTAG_FIELD_NUMBER: _ClassVar[int]
+    DATATYPE_FIELD_NUMBER: _ClassVar[int]
+    lex: str
+    langtag: str
+    datatype: int
+    def __init__(self, lex: _Optional[str] = ..., langtag: _Optional[str] = ..., datatype: _Optional[int] = ...) -> None: ...
+
+class RdfDefaultGraph(_message.Message):
+    __slots__ = ()
+    def __init__(self) -> None: ...
+
+class RdfTriple(_message.Message):
+    __slots__ = ("s_iri", "s_bnode", "s_literal", "s_triple_term", "p_iri", "p_bnode", "p_literal", "p_triple_term", "o_iri", "o_bnode", "o_literal", "o_triple_term")
+    S_IRI_FIELD_NUMBER: _ClassVar[int]
+    S_BNODE_FIELD_NUMBER: _ClassVar[int]
+    S_LITERAL_FIELD_NUMBER: _ClassVar[int]
+    S_TRIPLE_TERM_FIELD_NUMBER: _ClassVar[int]
+    P_IRI_FIELD_NUMBER: _ClassVar[int]
+    P_BNODE_FIELD_NUMBER: _ClassVar[int]
+    P_LITERAL_FIELD_NUMBER: _ClassVar[int]
+    P_TRIPLE_TERM_FIELD_NUMBER: _ClassVar[int]
+    O_IRI_FIELD_NUMBER: _ClassVar[int]
+    O_BNODE_FIELD_NUMBER: _ClassVar[int]
+    O_LITERAL_FIELD_NUMBER: _ClassVar[int]
+    O_TRIPLE_TERM_FIELD_NUMBER: _ClassVar[int]
+    s_iri: RdfIri
+    s_bnode: str
+    s_literal: RdfLiteral
+    s_triple_term: RdfTriple
+    p_iri: RdfIri
+    p_bnode: str
+    p_literal: RdfLiteral
+    p_triple_term: RdfTriple
+    o_iri: RdfIri
+    o_bnode: str
+    o_literal: RdfLiteral
+    o_triple_term: RdfTriple
+    def __init__(self, s_iri: _Optional[_Union[RdfIri, _Mapping]] = ..., s_bnode: _Optional[str] = ..., s_literal: _Optional[_Union[RdfLiteral, _Mapping]] = ..., s_triple_term: _Optional[_Union[RdfTriple, _Mapping]] = ..., p_iri: _Optional[_Union[RdfIri, _Mapping]] = ..., p_bnode: _Optional[str] = ..., p_literal: _Optional[_Union[RdfLiteral, _Mapping]] = ..., p_triple_term: _Optional[_Union[RdfTriple, _Mapping]] = ..., o_iri: _Optional[_Union[RdfIri, _Mapping]] = ..., o_bnode: _Optional[str] = ..., o_literal: _Optional[_Union[RdfLiteral, _Mapping]] = ..., o_triple_term: _Optional[_Union[RdfTriple, _Mapping]] = ...) -> None: ...
+
+class RdfQuad(_message.Message):
+    __slots__ = ("s_iri", "s_bnode", "s_literal", "s_triple_term", "p_iri", "p_bnode", "p_literal", "p_triple_term", "o_iri", "o_bnode", "o_literal", "o_triple_term", "g_iri", "g_bnode", "g_default_graph", "g_literal")
+    S_IRI_FIELD_NUMBER: _ClassVar[int]
+    S_BNODE_FIELD_NUMBER: _ClassVar[int]
+    S_LITERAL_FIELD_NUMBER: _ClassVar[int]
+    S_TRIPLE_TERM_FIELD_NUMBER: _ClassVar[int]
+    P_IRI_FIELD_NUMBER: _ClassVar[int]
+    P_BNODE_FIELD_NUMBER: _ClassVar[int]
+    P_LITERAL_FIELD_NUMBER: _ClassVar[int]
+    P_TRIPLE_TERM_FIELD_NUMBER: _ClassVar[int]
+    O_IRI_FIELD_NUMBER: _ClassVar[int]
+    O_BNODE_FIELD_NUMBER: _ClassVar[int]
+    O_LITERAL_FIELD_NUMBER: _ClassVar[int]
+    O_TRIPLE_TERM_FIELD_NUMBER: _ClassVar[int]
+    G_IRI_FIELD_NUMBER: _ClassVar[int]
+    G_BNODE_FIELD_NUMBER: _ClassVar[int]
+    G_DEFAULT_GRAPH_FIELD_NUMBER: _ClassVar[int]
+    G_LITERAL_FIELD_NUMBER: _ClassVar[int]
+    s_iri: RdfIri
+    s_bnode: str
+    s_literal: RdfLiteral
+    s_triple_term: RdfTriple
+    p_iri: RdfIri
+    p_bnode: str
+    p_literal: RdfLiteral
+    p_triple_term: RdfTriple
+    o_iri: RdfIri
+    o_bnode: str
+    o_literal: RdfLiteral
+    o_triple_term: RdfTriple
+    g_iri: RdfIri
+    g_bnode: str
+    g_default_graph: RdfDefaultGraph
+    g_literal: RdfLiteral
+    def __init__(self, s_iri: _Optional[_Union[RdfIri, _Mapping]] = ..., s_bnode: _Optional[str] = ..., s_literal: _Optional[_Union[RdfLiteral, _Mapping]] = ..., s_triple_term: _Optional[_Union[RdfTriple, _Mapping]] = ..., p_iri: _Optional[_Union[RdfIri, _Mapping]] = ..., p_bnode: _Optional[str] = ..., p_literal: _Optional[_Union[RdfLiteral, _Mapping]] = ..., p_triple_term: _Optional[_Union[RdfTriple, _Mapping]] = ..., o_iri: _Optional[_Union[RdfIri, _Mapping]] = ..., o_bnode: _Optional[str] = ..., o_literal: _Optional[_Union[RdfLiteral, _Mapping]] = ..., o_triple_term: _Optional[_Union[RdfTriple, _Mapping]] = ..., g_iri: _Optional[_Union[RdfIri, _Mapping]] = ..., g_bnode: _Optional[str] = ..., g_default_graph: _Optional[_Union[RdfDefaultGraph, _Mapping]] = ..., g_literal: _Optional[_Union[RdfLiteral, _Mapping]] = ...) -> None: ...
+
+class RdfGraphStart(_message.Message):
+    __slots__ = ("g_iri", "g_bnode", "g_default_graph", "g_literal")
+    G_IRI_FIELD_NUMBER: _ClassVar[int]
+    G_BNODE_FIELD_NUMBER: _ClassVar[int]
+    G_DEFAULT_GRAPH_FIELD_NUMBER: _ClassVar[int]
+    G_LITERAL_FIELD_NUMBER: _ClassVar[int]
+    g_iri: RdfIri
+    g_bnode: str
+    g_default_graph: RdfDefaultGraph
+    g_literal: RdfLiteral
+    def __init__(self, g_iri: _Optional[_Union[RdfIri, _Mapping]] = ..., g_bnode: _Optional[str] = ..., g_default_graph: _Optional[_Union[RdfDefaultGraph, _Mapping]] = ..., g_literal: _Optional[_Union[RdfLiteral, _Mapping]] = ...) -> None: ...
+
+class RdfGraphEnd(_message.Message):
+    __slots__ = ()
+    def __init__(self) -> None: ...
+
+class RdfNamespaceDeclaration(_message.Message):
+    __slots__ = ("name", "value")
+    NAME_FIELD_NUMBER: _ClassVar[int]
+    VALUE_FIELD_NUMBER: _ClassVar[int]
+    name: str
+    value: RdfIri
+    def __init__(self, name: _Optional[str] = ..., value: _Optional[_Union[RdfIri, _Mapping]] = ...) -> None: ...
+
+class RdfNameEntry(_message.Message):
+    __slots__ = ("id", "value")
+    ID_FIELD_NUMBER: _ClassVar[int]
+    VALUE_FIELD_NUMBER: _ClassVar[int]
+    id: int
+    value: str
+    def __init__(self, id: _Optional[int] = ..., value: _Optional[str] = ...) -> None: ...
+
+class RdfPrefixEntry(_message.Message):
+    __slots__ = ("id", "value")
+    ID_FIELD_NUMBER: _ClassVar[int]
+    VALUE_FIELD_NUMBER: _ClassVar[int]
+    id: int
+    value: str
+    def __init__(self, id: _Optional[int] = ..., value: _Optional[str] = ...) -> None: ...
+
+class RdfDatatypeEntry(_message.Message):
+    __slots__ = ("id", "value")
+    ID_FIELD_NUMBER: _ClassVar[int]
+    VALUE_FIELD_NUMBER: _ClassVar[int]
+    id: int
+    value: str
+    def __init__(self, id: _Optional[int] = ..., value: _Optional[str] = ...) -> None: ...
+
+class RdfStreamOptions(_message.Message):
+    __slots__ = ("stream_name", "physical_type", "generalized_statements", "rdf_star", "max_name_table_size", "max_prefix_table_size", "max_datatype_table_size", "logical_type", "version")
+    STREAM_NAME_FIELD_NUMBER: _ClassVar[int]
+    PHYSICAL_TYPE_FIELD_NUMBER: _ClassVar[int]
+    GENERALIZED_STATEMENTS_FIELD_NUMBER: _ClassVar[int]
+    RDF_STAR_FIELD_NUMBER: _ClassVar[int]
+    MAX_NAME_TABLE_SIZE_FIELD_NUMBER: _ClassVar[int]
+    MAX_PREFIX_TABLE_SIZE_FIELD_NUMBER: _ClassVar[int]
+    MAX_DATATYPE_TABLE_SIZE_FIELD_NUMBER: _ClassVar[int]
+    LOGICAL_TYPE_FIELD_NUMBER: _ClassVar[int]
+    VERSION_FIELD_NUMBER: _ClassVar[int]
+    stream_name: str
+    physical_type: PhysicalStreamType
+    generalized_statements: bool
+    rdf_star: bool
+    max_name_table_size: int
+    max_prefix_table_size: int
+    max_datatype_table_size: int
+    logical_type: LogicalStreamType
+    version: int
+    def __init__(self, stream_name: _Optional[str] = ..., physical_type: _Optional[_Union[PhysicalStreamType, str]] = ..., generalized_statements: bool = ..., rdf_star: bool = ..., max_name_table_size: _Optional[int] = ..., max_prefix_table_size: _Optional[int] = ..., max_datatype_table_size: _Optional[int] = ..., logical_type: _Optional[_Union[LogicalStreamType, str]] = ..., version: _Optional[int] = ...) -> None: ...
+
+class RdfStreamRow(_message.Message):
+    __slots__ = ("options", "triple", "quad", "graph_start", "graph_end", "namespace", "name", "prefix", "datatype")
+    OPTIONS_FIELD_NUMBER: _ClassVar[int]
+    TRIPLE_FIELD_NUMBER: _ClassVar[int]
+    QUAD_FIELD_NUMBER: _ClassVar[int]
+    GRAPH_START_FIELD_NUMBER: _ClassVar[int]
+    GRAPH_END_FIELD_NUMBER: _ClassVar[int]
+    NAMESPACE_FIELD_NUMBER: _ClassVar[int]
+    NAME_FIELD_NUMBER: _ClassVar[int]
+    PREFIX_FIELD_NUMBER: _ClassVar[int]
+    DATATYPE_FIELD_NUMBER: _ClassVar[int]
+    options: RdfStreamOptions
+    triple: RdfTriple
+    quad: RdfQuad
+    graph_start: RdfGraphStart
+    graph_end: RdfGraphEnd
+    namespace: RdfNamespaceDeclaration
+    name: RdfNameEntry
+    prefix: RdfPrefixEntry
+    datatype: RdfDatatypeEntry
+    def __init__(self, options: _Optional[_Union[RdfStreamOptions, _Mapping]] = ..., triple: _Optional[_Union[RdfTriple, _Mapping]] = ..., quad: _Optional[_Union[RdfQuad, _Mapping]] = ..., graph_start: _Optional[_Union[RdfGraphStart, _Mapping]] = ..., graph_end: _Optional[_Union[RdfGraphEnd, _Mapping]] = ..., namespace: _Optional[_Union[RdfNamespaceDeclaration, _Mapping]] = ..., name: _Optional[_Union[RdfNameEntry, _Mapping]] = ..., prefix: _Optional[_Union[RdfPrefixEntry, _Mapping]] = ..., datatype: _Optional[_Union[RdfDatatypeEntry, _Mapping]] = ...) -> None: ...
+
+class RdfStreamFrame(_message.Message):
+    __slots__ = ("rows", "metadata")
+    class MetadataEntry(_message.Message):
+        __slots__ = ("key", "value")
+        KEY_FIELD_NUMBER: _ClassVar[int]
+        VALUE_FIELD_NUMBER: _ClassVar[int]
+        key: str
+        value: bytes
+        def __init__(self, key: _Optional[str] = ..., value: _Optional[bytes] = ...) -> None: ...
+    ROWS_FIELD_NUMBER: _ClassVar[int]
+    METADATA_FIELD_NUMBER: _ClassVar[int]
+    rows: _containers.RepeatedCompositeFieldContainer[RdfStreamRow]
+    metadata: _containers.ScalarMap[str, bytes]
+    def __init__(self, rows: _Optional[_Iterable[_Union[RdfStreamRow, _Mapping]]] = ..., metadata: _Optional[_Mapping[str, bytes]] = ...) -> None: ...

pyjelly/options.py

@@ -0,0 +1,141 @@
+from __future__ import annotations
+
+import mimetypes
+from contextlib import suppress
+from dataclasses import dataclass
+from typing import Final
+from typing_extensions import Self
+
+from mypy_extensions import mypyc_attr
+
+from pyjelly import jelly
+from pyjelly.errors import (
+    JellyAssertionError,
+    JellyConformanceError,
+)
+
+MIN_NAME_LOOKUP_SIZE: Final[int] = 8
+
+MAX_LOOKUP_SIZE: Final[int] = 4096
+MIN_VERSION: Final[int] = 1
+MAX_VERSION: Final[int] = 2
+
+DEFAULT_NAME_LOOKUP_SIZE: Final[int] = 4000
+DEFAULT_PREFIX_LOOKUP_SIZE: Final[int] = 150
+DEFAULT_DATATYPE_LOOKUP_SIZE: Final[int] = 32
+
+STRING_DATATYPE_IRI = "http://www.w3.org/2001/XMLSchema#string"
+
+INTEGRATION_SIDE_EFFECTS: bool = True
+"""
+Whether to allow integration module imports to trigger side effects.
+
+These side effects are cheap and may include populating some registries
+for guessing the defaults for external integrations that work with Jelly.
+"""
+
+MIMETYPES = ("application/x-jelly-rdf",)
+
+
+def register_mimetypes(extension: str = ".jelly") -> None:
+    """
+    Associate files that have Jelly extension with Jelly MIME types.
+
+    >>> register_mimetypes()
+    >>> mimetypes.guess_type("out.jelly")
+    ('application/x-jelly-rdf', None)
+    """
+    for mimetype in MIMETYPES:
+        mimetypes.add_type(mimetype, extension)
+
+
+@mypyc_attr(allow_interpreted_subclasses=True)
+@dataclass(frozen=True)
+class LookupPreset:
+    max_names: int = DEFAULT_NAME_LOOKUP_SIZE
+    max_prefixes: int = DEFAULT_PREFIX_LOOKUP_SIZE
+    max_datatypes: int = DEFAULT_DATATYPE_LOOKUP_SIZE
+
+    def __post_init__(self) -> None:
+        if self.max_names < MIN_NAME_LOOKUP_SIZE:
+            msg = "name lookup size must be at least 8"
+            raise JellyConformanceError(msg)
+
+    @classmethod
+    def small(cls) -> Self:
+        return cls(max_names=128, max_prefixes=32, max_datatypes=32)
+
+
+@dataclass(frozen=True)
+class StreamTypes:
+    physical_type: jelly.PhysicalStreamType = jelly.PHYSICAL_STREAM_TYPE_UNSPECIFIED
+    logical_type: jelly.LogicalStreamType = jelly.LOGICAL_STREAM_TYPE_UNSPECIFIED
+
+    @property
+    def flat(self) -> bool:
+        return self.logical_type in (
+            jelly.LOGICAL_STREAM_TYPE_FLAT_TRIPLES,
+            jelly.LOGICAL_STREAM_TYPE_FLAT_QUADS,
+        )
+
+    def __repr__(self) -> str:
+        """
+        Return the representation of StreamTypes.
+
+        >>> repr(StreamTypes(9999, 8888))
+        'StreamTypes(9999, 8888)'
+        """
+        with suppress(ValueError):
+            physical_type_name = jelly.PhysicalStreamType.Name(self.physical_type)
+            logical_type_name = jelly.LogicalStreamType.Name(self.logical_type)
+            return f"StreamTypes({physical_type_name}, {logical_type_name})"
+        return f"StreamTypes({self.physical_type}, {self.logical_type})"
+
+    def __post_init__(self) -> None:
+        validate_type_compatibility(
+            physical_type=self.physical_type,
+            logical_type=self.logical_type,
+        )
+
+
+@dataclass(frozen=True)
+class StreamParameters:
+    generalized_statements: bool = False
+    rdf_star: bool = False
+    version: int = MAX_VERSION
+    delimited: bool = True
+    namespace_declarations: bool = False
+    stream_name: str = ""
+
+    def __post_init__(self) -> None:
+        selected = MAX_VERSION if self.namespace_declarations else MIN_VERSION
+        if not (MIN_VERSION <= selected <= MAX_VERSION):
+            msg = f"""Error occured while settin up the Stream options.
+            Version must be between {MIN_VERSION} and {MAX_VERSION}."""
+            raise JellyConformanceError(msg)
+        object.__setattr__(self, "version", selected)
+
+
+TRIPLES_ONLY_LOGICAL_TYPES = {
+    jelly.LOGICAL_STREAM_TYPE_GRAPHS,
+    jelly.LOGICAL_STREAM_TYPE_SUBJECT_GRAPHS,
+    jelly.LOGICAL_STREAM_TYPE_FLAT_TRIPLES,
+}
+
+
+def validate_type_compatibility(
+    physical_type: jelly.PhysicalStreamType,
+    logical_type: jelly.LogicalStreamType,
+) -> None:
+    if (
+        physical_type == jelly.PHYSICAL_STREAM_TYPE_UNSPECIFIED
+        or logical_type == jelly.LOGICAL_STREAM_TYPE_UNSPECIFIED
+    ):
+        return
+    triples_physical_type = physical_type == jelly.PHYSICAL_STREAM_TYPE_TRIPLES
+    triples_logical_type = logical_type in TRIPLES_ONLY_LOGICAL_TYPES
+    if triples_physical_type != triples_logical_type:
+        physical_type_name = jelly.PhysicalStreamType.Name(physical_type)
+        logical_type_name = jelly.LogicalStreamType.Name(logical_type)
+        msg = f"{physical_type_name} is not compatible with {logical_type_name}"
+        raise JellyAssertionError(msg)