def main():
from rita.utils import RitaJSONEncoder
parser = argparse.ArgumentParser(
description="Compile rita -> spaCy patterns"
)
parser.add_argument("-f", help=".rita rules file")
parser.add_argument(
"out",
help="output .jsonl file to store rules"
)
parser.add_argument("--debug", help="debug mode", action="store_true")
parser.add_argument("--engine", help="Engine to use when compiling rules", default="spacy")
args = parser.parse_args()
if args.debug:
logging.basicConfig(level=logging.DEBUG)
else:
logging.basicConfig(level=logging.INFO)
patterns = rita.compile(args.f, use_engine=args.engine)
logger.info("Compiling rules using {} engine".format(args.engine))
with open(args.out, "w") as f:
for pattern in patterns:
f.write(json.dumps(pattern, cls=RitaJSONEncoder) + "\n")
def _spacy_v3(model, patterns=None, rules_path=None, rules_string=None, override_ents=True):
ruler = model.add_pipe("entity_ruler", config={"overwrite_ents": override_ents, "validate": True})
if not patterns:
if rules_path:
patterns = rita.compile(rules_path, use_engine="spacy")
elif rules_string:
patterns = rita.compile_string(rules_string, use_engine="spacy")
else:
raise RuntimeError("Please provides rules. Either `patterns`, `rules_path` or `rules_string`")
ruler.add_patterns(patterns)
else:
ruler.from_disk(patterns)
return model
def test_shortcuts_spacy_compiled():
spacy = pytest.importorskip("spacy", minversion="2.1")
nlp = spacy.load("en_core_web_sm")
tmp = tempfile.NamedTemporaryFile(mode="w",
encoding="UTF-8",
suffix=".jsonl",
delete=False)
patterns = rita.compile("examples/color-car.rita")
for pattern in patterns:
tmp.write(json.dumps(pattern) + "\n")
tmp.flush()
tmp.close()
setup_spacy(nlp, patterns=tmp.name)
os.unlink(tmp.name)
def _spacy_v2(model, patterns=None, rules_path=None, rules_string=None, override_ents=True):
from spacy.pipeline import EntityRuler
ruler = EntityRuler(model, overwrite_ents=override_ents)
if not patterns:
if rules_path:
patterns = rita.compile(rules_path, use_engine="spacy")
elif rules_string:
patterns = rita.compile_string(rules_string, use_engine="spacy")
else:
raise RuntimeError("Please provides rules. Either `patterns`, `rules_path` or `rules_string`")
ruler.add_patterns(patterns)
else:
ruler.from_disk(patterns)
model.add_pipe(ruler)
return model