Python library for making calls to SciBite's NER engine, TERMite, as well as the TExpress module for defining more complex semantic patterns. The library also enables post-processing of the JSON returned from such requests.
$ pip3 install termite_toolkit==[version]
Versions listed on PyPi!
from termite_toolkit import termite
from pprint import pprint
# specify termite API endpoint
termite_home = "http://localhost:9090/termite"
# specify entities to annotate
entities = "DRUG"
# initialise a request builder
t = termite.TermiteRequestBuilder()
# individually add items to your TERMite request
t.set_url(termite_home)
t.set_fuzzy(True)
t.set_text("citrate macrophage colony sildenafil stimulating factor influenza hedgehog")
t.set_entities(entities)
t.set_subsume(True)
t.set_input_format("txt")
t.set_output_format("doc.jsonx")
t.set_reject_ambiguous(False)
t.set_options({'fragmentSize': 20})
# execute the request
termite_response = t.execute(display_request=True)
pprint(termite_response)
from pprint import pprint
from termite_toolkit import texpress
# specify termite API endpoint
termite_home = "http://localhost:9090/termite"
# specify the pattern you wish to search for- this can created in the TERMite UI
pattern = ":(INDICATION):{0,5}:(GENE)"
t = texpress.TexpressRequestBuilder()
# individually add items to your TERMite request
t.set_url(termite_home)
t.set_text("breast cancer brca1")
t.set_subsume(True)
t.set_input_format("txt")
t.set_output_format("json")
t.set_allow_ambiguous(False)
t.set_pattern(pattern)
# execute the request
texpress_response = t.execute(display_request=True)
pprint(texpress_response)
Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.