def _add_custom_spacy_extensions(self):
for n, f in self.is_attrs_name2func:
ext = Token.get_extension(n)
if ext is None:
Token.set_extension(n, getter=f, force=True)
for item in [Span, Doc]:
for n, f in self.has_attrs_name2func:
ext = item.get_extension(n)
if ext is None:
#print(f"Setting: {item}.set_extension({n}, getter= {f})")
item.set_extension(n, getter=f, force=True)
# Add Attr Getters for Span (i.e. Doc.ents)
for n, f in self.get_attrs_name2func:
ext = Span.get_extension(n)
if ext is None:
Span.set_extension(n, getter=f, force=True)
def __init__(self,
nlp: Language,
attr: str = "spaczz_ent",
**cfg: Any) -> None:
"""Initialize the spaczz ruler with a Language object and cfg parameters.
All spaczz ruler cfg parameters are prepended with "spaczz_".
If spaczz_patterns is supplied here, they need to be a list of spaczz patterns:
dictionaries with a "label", "pattern", "type", and optional "kwargs" key.
For example:
{'label': 'ORG', 'pattern': 'Apple', 'type': 'fuzzy', 'kwargs': {'min_r2': 90}}.
Args:
nlp: The shared nlp object to pass the vocab to the matchers
(not currently used by spaczz matchers) and process fuzzy patterns.
attr: Name of custom Span attribute that denotes whether an
entity was added via the spaczz ruler or not.
Default is "spaczz_ent".
**cfg: Other config parameters. The SpaczzRuler makes heavy use
of cfg to pass additional parameters down to the matchers.
spaczz config parameters start with "spaczz_" to keep them
from colliding with other cfg components.
SpaczzRuler cfg components include (with "spaczz_" prepended to them):
overwrite_ents (bool): Whether to overwrite exisiting Doc.ents
with new matches. Default is False.
ent_id_sep (str): String to separate entity labels and ids on.
regex_config (Union[str, RegexConfig]): Config to use with the
regex matcher. Default is "default". See RegexMatcher/RegexSearcher
documentation for available parameter details.
fuzzy_defaults (Dict[str, Any]): Modified default parameters to use with
the fuzzy matcher. Default is an empty dictionary -
utilizing defaults.
regex_defaults (Dict[str, Any]): Modified default parameters to use with
the regex matcher. Default is an empty dictionary -
utilizing defaults. See RegexMatcher/RegexSearcher documentation
for parameter details.
patterns (Iterable[Dict[str, Any]]): Patterns to initialize
the ruler with. Default is None.
If SpaczzRuler is loaded as part of a model pipeline,
cfg will include all keyword arguments passed to spacy.load.
Raises:
TypeError: If spaczz_{name}_defaults passed are not dictionaries.
"""
if not Span.get_extension(attr):
Span.set_extension(attr, default=False)
self.nlp = nlp
self.fuzzy_patterns: DefaultDict[str, DefaultDict[
str, Any]] = defaultdict(lambda: defaultdict(list))
self.regex_patterns: DefaultDict[str, DefaultDict[
str, Any]] = defaultdict(lambda: defaultdict(list))
self.ent_id_sep = cfg.get("spaczz_ent_id_sep", DEFAULT_ENT_ID_SEP)
self._ent_ids: Dict[Any, Any] = defaultdict(dict)
self.overwrite = cfg.get("spaczz_overwrite_ents", False)
default_names = ("spaczz_fuzzy_defaults", "spaczz_regex_defaults")
self.defaults = {}
for name in default_names:
if name in cfg:
if isinstance(cfg[name], dict):
self.defaults[name] = cfg[name]
else:
raise TypeError((
"Defaults must be a dictionary of keyword arguments,",
f"not {type(cfg[name])}.",
))
self.fuzzy_matcher = FuzzyMatcher(
nlp.vocab,
**self.defaults.get("spaczz_fuzzy_defaults", {}),
)
self.regex_matcher = RegexMatcher(
nlp.vocab,
cfg.get("spaczz_regex_config", "default"),
**self.defaults.get("spaczz_regex_defaults", {}),
)
patterns = cfg.get("spaczz_patterns")
if patterns is not None:
self.add_patterns(patterns)
