Source code

Download source archive ↓

src/rdflib/plugins/serializers/nt.py

1from __future__ import annotations
2
3import codecs
4import warnings
5from typing import IO, TYPE_CHECKING, Any, Optional, Tuple, Union
6
7from rdflib.graph import Graph
8from rdflib.serializer import Serializer
9from rdflib.term import Literal
10
11if TYPE_CHECKING:
12    from rdflib.graph import _TripleType
13
14"""
15N-Triples RDF graph serializer for RDFLib.
16See <http://www.w3.org/TR/rdf-testcases/#ntriples> for details about the
17format.
18"""
19
20__all__ = ["NTSerializer"]
21
22
23class NTSerializer(Serializer):
24    """Serializes RDF graphs to NTriples format."""
25
26    def __init__(self, store: Graph):
27        Serializer.__init__(self, store)
28
29    def serialize(
30        self,
31        stream: IO[bytes],
32        base: Optional[str] = None,
33        encoding: Optional[str] = "utf-8",
34        **kwargs: Any,
35    ) -> None:
36        if base is not None:
37            warnings.warn("NTSerializer does not support base.")
38        if encoding != "utf-8":
39            warnings.warn(
40                "NTSerializer always uses UTF-8 encoding. "
41                f"Given encoding was: {encoding}"
42            )
43
44        for triple in self.store:
45            stream.write(_nt_row(triple).encode())
46
47
48class NT11Serializer(NTSerializer):
49    """Serializes RDF graphs to RDF 1.1 NTriples format.
50
51    Exactly like nt - only utf8 encoded.
52    """
53
54    def __init__(self, store: Graph):
55        Serializer.__init__(self, store)  # default to utf-8
56
57
58def _nt_row(triple: _TripleType) -> str:
59    if isinstance(triple[2], Literal):
60        return "%s %s %s .\n" % (
61            triple[0].n3(),
62            triple[1].n3(),
63            _quoteLiteral(triple[2]),
64        )
65    else:
66        return "%s %s %s .\n" % (triple[0].n3(), triple[1].n3(), triple[2].n3())
67
68
69def _quoteLiteral(l_: Literal) -> str:  # noqa: N802
70    """A simpler version of term.Literal.n3()"""
71
72    encoded = _quote_encode(l_)
73
74    if l_.language:
75        if l_.datatype:
76            raise Exception("Literal has datatype AND language!")
77        return "%s@%s" % (encoded, l_.language)
78    elif l_.datatype:
79        return "%s^^<%s>" % (encoded, l_.datatype)
80    else:
81        return "%s" % encoded
82
83
84def _quote_encode(l_: str) -> str:
85    return '"%s"' % l_.replace("\\", "\\\\").replace("\n", "\\n").replace(
86        '"', '\\"'
87    ).replace("\r", "\\r")
88
89
90def _nt_unicode_error_resolver(
91    err: UnicodeError,
92) -> Tuple[Union[str, bytes], int]:
93    """
94    Do unicode char replaces as defined in https://www.w3.org/TR/2004/REC-rdf-testcases-20040210/#ntrip_strings
95    """
96
97    def _replace_single(c):
98        c = ord(c)
99        fmt = "\\u%04X" if c <= 0xFFFF else "\\U%08X"
100        return fmt % c
101
102    # type error: "UnicodeError" has no attribute "object"
103    # type error: "UnicodeError" has no attribute "start"
104    # type error: "UnicodeError" has no attribute "end"
105    string = err.object[err.start : err.end]  # type: ignore[attr-defined]
106    # type error: "UnicodeError" has no attribute "end"
107    return "".join(_replace_single(c) for c in string), err.end  # type: ignore[attr-defined]
108
109
110codecs.register_error("_rdflib_nt_escape", _nt_unicode_error_resolver)