def map_object_literal(
v: Union[str, jsonasobj.JsonObj]) -> ShExJ.ObjectLiteral:
""" `PyShEx.jsg <https://github.com/hsolbrig/ShExJSG/ShExJSG/ShExJ.jsg>`_ does not add identifying
types to ObjectLiterals. This routine re-identifies the types
"""
# TODO: isinstance(v, JSGString) should work here, but it doesn't with IRIREF(http://a.example/v1)
return v if issubclass(type(v), JSGString) or (isinstance(v, JSGObject) and 'type' in v) else \
ShExJ.IRIREF(v) if isinstance(v, str) else ShExJ.ObjectLiteral(**v._as_dict)
def shapeRef_to_iriref(self,
ref: ShExDocParser.ShapeRefContext) -> ShExJ.IRIREF:
""" shapeRef: ATPNAME_NS | ATPNAME_LN | '@' shapeExprLabel """
if ref.ATPNAME_NS():
return ShExJ.IRIREF(
self._lookup_prefix(ref.ATPNAME_NS().getText()[1:]))
elif ref.ATPNAME_LN():
prefix, local = ref.ATPNAME_LN().getText()[1:].split(':', 1)
return ShExJ.IRIREF(
self._lookup_prefix(prefix + ':') + (local if local else ""))
else:
return self.shapeexprlabel_to_IRI(ref.shapeExprLabel())
def tripleexprlabel_to_iriref(self, tripleExprLabel: ShExDocParser.TripleExprLabelContext) \
-> Union[ShExJ.BNODE, ShExJ.IRIREF]:
""" tripleExprLabel: iri | blankNode """
if tripleExprLabel.iri():
return self.iri_to_iriref(tripleExprLabel.iri())
else:
return ShExJ.BNODE(tripleExprLabel.blankNode().getText())
def prefixedname_to_iriref(
self, prefix: ShExDocParser.PrefixedNameContext) -> ShExJ.IRIREF:
""" prefixedName: PNAME_LN | PNAME_NS
PNAME_NS: PN_PREFIX? ':' ;
PNAME_LN: PNAME_NS PN_LOCAL ;
"""
return ShExJ.IRIREF(self.prefixedname_to_str(prefix))
def shapeexprlabel_to_IRI(self, shapeExprLabel: ShExDocParser.ShapeExprLabelContext) \
-> Union[ShExJ.BNODE, ShExJ.IRIREF]:
""" shapeExprLabel: iri | blankNode """
if shapeExprLabel.iri():
return self.iri_to_iriref(shapeExprLabel.iri())
else:
return ShExJ.BNODE(shapeExprLabel.blankNode().getText())
def test_predicate_scan(self):
c = setup_context(shex_1, rdf_1)
predicates: List[URIRef] = []
c.visit_shapes(
c.shapeExprFor(IRIREF('http://schema.example/UserShape')),
triple_expr_finder, predicates)
self.assertEqual([RDF.type], predicates)
# Quick test of the utility function
self.assertEqual(
predicates_in_expression(
c.shapeExprFor(IRIREF('http://schema.example/UserShape')), c),
[ShExJ.IRIREF(str(u)) for u in predicates])
def literal_to_ObjectLiteral(
self,
literal: ShExDocParser.LiteralContext) -> ShExJ.ObjectLiteral:
""" literal: rdfLiteral | numericLiteral | booleanLiteral """
rval = ShExJ.ObjectLiteral()
if literal.rdfLiteral():
rdflit = literal.rdfLiteral()
txt = rdflit.string().getText()
quote_char = ''
if len(txt) > 5 and (txt.startswith("'''") and txt.endswith("'''")
or txt.startswith('"""')
and txt.endswith('"""')):
txt = txt[3:-3]
else:
quote_char = txt[0]
txt = txt[1:-1]
txt = self.fix_text_escapes(txt, quote_char)
rval.value = String(txt)
if rdflit.LANGTAG():
rval.language = ShExJ.LANGTAG(
rdflit.LANGTAG().getText()[1:].lower())
if rdflit.datatype():
rval.type = self.iri_to_str(rdflit.datatype().iri())
elif literal.numericLiteral():
numlit = literal.numericLiteral()
if numlit.INTEGER():
rval.value = String(numlit.INTEGER().getText())
rval.type = RDF_INTEGER_TYPE
elif numlit.DECIMAL():
rval.value = String(numlit.DECIMAL().getText())
rval.type = RDF_DECIMAL_TYPE
elif numlit.DOUBLE():
rval.value = String(numlit.DOUBLE().getText())
rval.type = RDF_DOUBLE_TYPE
elif literal.booleanLiteral():
rval.value = String(literal.booleanLiteral().getText().lower())
rval.type = RDF_BOOL_TYPE
return rval
def iri_to_iriref(self, iri_: ShExDocParser.IriContext) -> ShExJ.IRIREF:
""" iri: IRIREF | prefixedName
prefixedName: PNAME_LN | PNAME_NS
"""
return ShExJ.IRIREF(self.iri_to_str(iri_))
def iriref_to_shexj_iriref(self,
ref: ShExDocParser.IRIREF) -> ShExJ.IRIREF:
""" IRIREF: '<' (~[\u0000-\u0020=<>\"{}|^`\\] | UCHAR)* '>'
IRI: (PN_CHARS | '!' | ''.' | ':' | '/' | '\\' | '#' | '@' | '%' | '&' | UCHAR)* """
return ShExJ.IRIREF(self.iriref_to_str(ref))
def __init__(self):
self.schema = ShExJ.Schema()
self.ld_prefixes: Dict[PREFIXstr,
IRIstr] = {} # prefixes in the JSON-LD module
self.prefixes: Dict[PREFIXstr, IRIstr] = {} # Assigned prefixes
self.base: Optional[IRIstr] = None
import re
from typing import Optional
from ShExJSG import ShExJ
from pyjsg.jsglib import *
from rdflib import RDF, XSD
from pyshexc.parser.ShExDocParser import ShExDocParser
IRIstr = str
PREFIXstr = str
RDF_TYPE = ShExJ.IRIREF(str(RDF.type))
RDF_INTEGER_TYPE = ShExJ.IRIREF(str(XSD.integer))
RDF_DOUBLE_TYPE = ShExJ.IRIREF(str(XSD.double))
RDF_DECIMAL_TYPE = ShExJ.IRIREF(str(XSD.decimal))
RDF_BOOL_TYPE = ShExJ.IRIREF(str(XSD.boolean))
class ParserContext:
"""
Context maintained across ShExC parser implementation modules
"""
def __init__(self):
self.schema = ShExJ.Schema()
self.ld_prefixes: Dict[PREFIXstr,
IRIstr] = {} # prefixes in the JSON-LD module
self.prefixes: Dict[PREFIXstr, IRIstr] = {} # Assigned prefixes
self.base: Optional[IRIstr] = None
def _resolve_relative_uri(
self, ref: Union[URIRef, BNode,
ShExJ.shapeExprLabel]) -> ShExJ.shapeExprLabel:
return ShExJ.IRIREF(str(self.base_namespace[str(
ref)])) if ':' not in str(ref) and self.base_namespace else ref
def eval_entry(self, entry_name: str) -> bool:
mes = self.mfst.entries[entry_name]
for me in mes: # There can be more than one entry per name...
# Determine the start point
if not self.started:
if not me.name.startswith(START_AFTER):
self.start_skipped += 1
return True
else:
self.started = True
if VERBOSE:
print(
f"STARTED - Skipped {self.start_skipped} entries")
# Determine whether this entry should be skipped
should_skip = False
# Skip
skipped_traits = list(me.traits.intersection(skip_traits))
if skipped_traits:
if VERBOSE:
print(
f"Skipping {me.name} ({', '.join([self.URIname(t) for t in me.traits])}) - Skipped trait"
)
key = str(skipped_traits[0]).replace(str(SHT), 'sht:')
if key not in self.skip_reasons:
self.skip_reasons[key] = 0
self.skip_reasons[key] = self.skip_reasons[key] + 1
self.skip(me.name)
should_skip = True
elif me.name in self.expected_failures:
if VERBOSE:
print(
f"Skipping {me.name} ({', '.join([self.URIname(t) for t in me.traits])})"
f" - {self.expected_failures[me.name]}")
key = self.expected_failures[me.name]
if key not in self.skip_reasons:
self.skip_reasons[key] = 0
self.skip_reasons[key] = self.skip_reasons[key] + 1
self.skip(me.name)
should_skip = True
if should_skip and not TEST_SKIPS_ONLY:
return True
if TEST_SKIPS_ONLY and not should_skip:
return True
# Validate the entry
if VERBOSE:
shex_uri = self.mfst.schema_loader.location_rewrite(
me.schema_uri)
data_uri = self.mfst.data_redirector.uri_for(me.data_uri) \
if self.mfst.data_redirector else me.data_uri
print(
f"Testing {me.name} ({'P' if me.should_pass else 'F'}): {shex_uri} - {data_uri}"
)
g, s = me.data_graph(), me.shex_schema()
if g is None and me.data_uri:
print("\t ERROR: Unable to load data file")
print(f"\t TRAITS: ({','.join(me.traits)})")
self.skip(me.name)
return True
if not s:
print(f"\t ERROR: Unable to load schema {me.schema_uri}")
print(f"\t TRAITS: ({','.join(me.traits)})")
self.nskipped += 1
self.skip(me.name)
return False
cntxt = Context(g,
s,
me.extern_shape_for,
base_namespace=BASE_FILE_LOC)
cntxt.debug_context.debug = DEBUG
map_ = FixedShapeMap()
focus = self.mfst.data_uri(me.focus)
if not focus:
print("\t***** FAIL *****")
print(f"\tFocus: {me.focus} not in schema")
print(f"\t TRAITS: ({','.join(me.traits)})")
self.fail(me.name)
return False
# if ':' not in focus:
# focus = "file://" + focus
map_.add(
ShapeAssociation(
focus,
ShExJ.IRIREF(me.shape) if me.shape else START))
#################################
# Actual validation occurs here
#################################
rslt = isValid(cntxt, map_)
test_result, reasons = rslt[0] or not me.should_pass, rslt[1]
# Analyze the result
if not VERBOSE and not test_result:
print(
f"Failed {me.name} ({'P' if me.should_pass else 'F'}): {me.schema_uri} - {me.data_uri}"
)
print(f"\t TRAITS: ({','.join(me.traits)})")
if test_result:
self.pass_(me.name)
else:
if VERBOSE:
print("\t**** FAIL *****")
print(f"\t TRAITS: ({','.join(me.traits)})")
for reason in reasons:
print(f"\t{reason}")
self.fail(me.name)
return test_result
def test_extern_str(self):
mfst = ShExManifest(manifest_ttl, 'turtle')
me = mfst.entries['shapeExtern_pass'][0]
self.assertIsNotNone(
me.extern_shape_for(ShExJ.IRIREF("http://a.example/Sext")))
def __init__(self):
self.schema = ShExJ.Schema()
self.ld_prefixes = {} # prefixes in the JSON-LD module
self.prefixes = {} # Assigned prefixes
self.base = None