Spaces:
Sleeping
Sleeping
| import subprocess | |
| import spacy | |
| import pyinflect | |
| import pandas as pd | |
| from typing import Tuple, Union, Any | |
| class Parser: | |
| def __init__( | |
| self | |
| ) -> None: | |
| self.parser = self.__init_parser("en_core_web_sm") | |
| self.__overrides = self.__load_overrides("data/overrides.csv") | |
| def __load_overrides( | |
| self, | |
| file_path: str | |
| ) -> Any: | |
| dic = pd.read_csv(file_path).set_index('VB').to_dict() | |
| fun = lambda verb, tag: dic[tag].get(verb, False) | |
| return fun | |
| def __init_parser( | |
| self, | |
| model: str | |
| ) -> spacy.language: | |
| parser = None | |
| try: | |
| parser = spacy.load(model) | |
| except: | |
| print(f"* Downloading {model} model...") | |
| _ = subprocess.Popen( | |
| f"python -m spacy download {model}", | |
| stdout=subprocess.PIPE, | |
| shell=True | |
| ).communicate() | |
| parser = spacy.load(model) | |
| return parser | |
| def __format_error( | |
| self, | |
| error: str | |
| ) -> str: | |
| template = """ | |
| <center> | |
| <div class="alert alert-warning" role="alert"> | |
| <h6><b>{}</b></h6> | |
| </div> | |
| </center> | |
| """ | |
| return template.format(error) | |
| def __format_output( | |
| self, | |
| infinitive: str, | |
| past: str, | |
| participle: str, | |
| ) -> str: | |
| template = """ | |
| |Infinitive| Simple Past | Past Participle | | |
| | :----: | :----: | :----: | | |
| |{} | {}| {}| | |
| """ | |
| return template.format(infinitive, past, participle) | |
| def is_in_overrides( | |
| self, | |
| verb: str, | |
| tense: str | |
| ) -> Tuple[str, bool]: | |
| res = self.__overrides(verb, tense) | |
| if isinstance(res, bool): | |
| return res, "" | |
| return True, res | |
| def __get_inflections( | |
| self, | |
| infinitive: spacy.tokens.token.Token, | |
| tense: str | |
| ) -> Union[str, None]: | |
| # Check if verb is in overrides file | |
| res, inflections = self.is_in_overrides(infinitive.text, tense) | |
| if not res: | |
| form1 = infinitive._.inflect(tense, form_num=0) | |
| form2 = infinitive._.inflect(tense, form_num=1) | |
| inflections = list(set([form1, form2])) | |
| if len(inflections) == 1 and inflections[0] == None: | |
| return None | |
| inflections = '/'.join(inflections) | |
| return inflections | |
| def __getAllTenses( | |
| self, | |
| tk_verb: spacy.tokens.token.Token | |
| ) -> Tuple[str,str,str,str]: | |
| infinitive = tk_verb._.inflect('VB') | |
| past = self.__get_inflections(tk_verb, 'VBD') | |
| participle = self.__get_inflections(tk_verb, 'VBN') | |
| return infinitive, past, participle | |
| def __tokenizer( | |
| self, | |
| verb: str | |
| ) -> spacy.tokens.token.Token: | |
| return self.parser(verb)[0] | |
| def get( | |
| self, | |
| verb: str | |
| ) -> Tuple[str,str]: | |
| verb = verb.strip().lower() | |
| error, output = "", "" | |
| if verb == "": | |
| error = self.__format_error( | |
| f"Error: The Verb field can not be empty!" | |
| ) | |
| return error, output | |
| tk_verb = self.__tokenizer(verb) | |
| infinitive, past, participle = self.__getAllTenses(tk_verb) | |
| if infinitive is None or past is None or participle is None: | |
| error = self.__format_error( | |
| f"Error: The verb '<b>{verb}</b>' has not been found or not spelled correctly!" | |
| ) | |
| return error, output | |
| print(f"{verb} -> {infinitive},{past},{participle}") | |
| output = self.__format_output(infinitive, past, participle) | |
| return error, output |