phrase-detective 0.1.15

Phrase recognizer component for spacy pipeline

Installation from pip3

pip3 install --verbose phrase_detective 
python -m spacy download en_core_web_trf
python -m spacy download es_dep_news_trf
python -m spacy download de_dep_news_trf

Usage

Please refer to api docs

Detect noun phrases

import spacy
from spacy import Language
from phrase_detective import NounPhraseRecognizer, PKG_INDICES

@Language.factory("nprecog")
def create_np_parser(nlp: Language, name: str):
  return NounPhraseRecognizer(nlp) 

def noun_phrase(lang, sentence):
  nlp = spacy.load(PKG_INDICES[lang])
  nlp.add_pipe("nprecog")
  doc = nlp(sentence)
  for np in doc._.noun_phrases:
    print(np.text)

Detect preposition phrases

import spacy
from spacy import Language
from phrase_detective import PrepPhraseRecognizer, PKG_INDICES

@Language.factory("pprecog")
def create_pp_parser(nlp: Language, name: str):
  return PrepPhraseRecognizer(nlp) 

def prep_phrase(lang, sentence):
  nlp = spacy.load(PKG_INDICES[lang])
  nlp.add_pipe("pprecog")
  doc = nlp(sentence)
  for np in doc._.prep_phrases:
    print(np.text)

Detect verb phrases

import spacy
from spacy import Language
from phrase_detective import VerbKnowledgeRecognizer, PKG_INDICES

@Language.factory("vkbrecog")
def create_vkb_parser(nlp: Language, name: str):
  return VerbKnowledgeRecognizer(nlp) 

def verb_knowledge(lang, sentence):
  nlp = spacy.load(PKG_INDICES[lang])
  nlp.add_pipe("vkbrecog")
  doc = nlp(sentence)
  for v in doc._.verbs:
    print("TEXT: {}, TAG: {}, FORM: {}, ORIGNAL: {}".format(v.text, v.tag_, spacy.explain(v.tag_), v.lemma_))
  for pp in doc._.passive_phrases:
    print(pp.text)
  for vp in doc._.verb_phrases:
    print(vp)

Development

Clone project

git clone https://github.com/qishe-nlp/phrase-detective.git

Install poetry

Install dependencies

poetry update

Test and Issue

poetry run pytest -rP

which run tests under tests/*

Create sphix docs

poetry shell
cd apidocs
sphinx-apidoc -f -o source ../phrase_detective
make html
python -m http.server -d build/html

Hose docs on github pages

cp -rf apidocs/build/html/* docs/

Build

• Change version in pyproject.toml and phrase_detective/__init__.py
• Build python package by poetry build

Git commit and push

Publish from local dev env

• Set pypi test environment variables in poetry, refer to poetry doc
• Publish to pypi test by poetry publish -r test

Publish through CI

• Github action build and publish package to test pypi repo

git tag [x.x.x]
git push origin master

• Manually publish to pypi repo through github action