main.py

from countriesIdentification import identify_locations
from datesIdentification import dates_binding
from magnitudeIdentification import magnitude_binding
from comparativesIdentification import comparatives_binding
from earthquaqeIdentification import identify_earthquake_event


def process_final_dict(final_dictionary):
  """
  Function to convert each one of the error codes from each component into a relevant code number to be handled by the SF
  """

  # convert all tuple error messages into dictionary error messages
  for i, elem in enumerate(final_dictionary):
    if isinstance(elem, tuple):

      if elem == (0, "MAGNITUDE", "no_magnitude"):
        final_dictionary[i] = {"Number": 9999911}

      elif elem == (0, "MAGNITUDE", "more_magnitude"):
        final_dictionary[i] = {"Number": 9999912}

      elif elem == (0, "MAGNITUDE", "format_error"):
        final_dictionary[i] = {"Number": 9999914}

      elif elem == (0, "MAGNITUDE", "unknown_error"):
        final_dictionary[i] = {"Number": 9999913}

      elif elem == (0, "EARTHQUAKE_EVENT", "no_earthquake_reference"):
        final_dictionary[i] = {"event":9999921}

      elif elem == (0, "EARTHQUAKE_EVENT", "unknown_error"):
        final_dictionary[i] = {"event": 9999922}

      elif elem == (0,'DATES', 'wrong_date_format'):
        final_dictionary[i] = {"date": {"day": 9999931, "month": 9999931, "year": 9999931}}

      elif elem == (0,'DATES', 'no_date'):
        final_dictionary[i] = {"date": {"day": 9999932, "month": 9999932, "year": 9999932}}

      elif elem == (0,'DATES', 'more_dates'):
        final_dictionary[i] = {"date": {"day": 9999933, "month": 9999933, "year": 9999933}}

      elif elem == (0,'DATES', 'unknown_error'):
        final_dictionary[i] = {"date": {"day": 9999934, "month": 9999934, "year": 9999934}}

      elif elem == (0, "LOCATION", "no_country"):
        final_dictionary[i] = {"city":[9999941], "country":[9999941]}

      elif elem == (0, "LOCATION", "more_city_or_country"):
        final_dictionary[i] = {"city": [9999942], "country": [9999942]}

      elif elem == (0, "LOCATION", "more_country"):
        final_dictionary[i] = {"city": [9999943], "country": [9999943]}

      elif elem == (0, "LOCATION", "unknown_error"):
        final_dictionary[i] = {"city": [9999944], "country": [9999944]}

      elif elem == (0, "COMPARATIVES", "more_comparatives_mentions"):
        final_dictionary[i] = {"comparative": 9999951}

      elif elem == (0, "COMPARATIVES", "no_comparatives"):
        final_dictionary[i] = {"comparative": 9999952}

      elif elem == (0, "COMPARATIVES", "more_symbol_comparatives"):
        final_dictionary[i] = {"comparative": 9999953}

      elif elem == (0, "COMPARATIVES", "unknown_error"):
        final_dictionary[i] = {"comparative": 9999954}

  return final_dictionary



def natural_language_module(sentence):
  """
  Function to execute the complete natural language module pipeline
  """
  try:
    final_dictionary = []

    # identify whether the sentence is referred on earthquake events
    earth_event = identify_earthquake_event(sentence)

    if earth_event:
      final_dictionary.append(earth_event)

    # identify the target country and city in the sentence
    location = identify_locations(sentence)

    if location:
      final_dictionary.append(location)

    # identify the target comparative in the sentence
    comparative = comparatives_binding(sentence)

    if comparative:
      final_dictionary.append(comparative)

    # identify the target date in the sentence
    date = dates_binding(sentence)

    if isinstance(date, list):

        date_dict = date[0]
        date_replc = date[1]

        if date_dict:
          final_dictionary.append(date_dict[0])

          # we also delete the date reference from the sentence so that there will
          # not be any confusion with it for the magnitude identification module
          if len(date_replc) == 1:
            sentence = sentence.replace(date_replc[0], " ")

    # in case it is a tuple we add it as it is and we do not substitute something in the sentence
    elif isinstance(date, tuple):
      final_dictionary.append(date)

    # identify the target magnitude number in the sentence
    magnitude = magnitude_binding(sentence)

    if magnitude:
      final_dictionary.append(magnitude)

    clean_final_dictionary = process_final_dict(final_dictionary)

    result = {}
    for d in clean_final_dictionary:
      result.update(d)

    return result

  except:
    return "\n\n=== AN UNEXPECTED ERROR HAS OCCURED. PLEASE EXECUTE AGAIN THE SCRIPT OR COMMUNICATE WITH THE DEVELOPER TEAM === \n\n"


def replace_zero_with_null(d):
  """
  This is a small helper function to convert the 0 references on the final json to be sent, into None, as needed by the SF
  """
  for k, v in d.items():
    if isinstance(v, dict):
      replace_zero_with_null(v)
    elif v == 0 and k != "point":
      d[k] = "null"
  return d


def process_json_sf(nl_json, sentence):
  """
  Function to convert the captured information an a relevant json format
  """
  try:

    sf_json_format = {
      "text": sentence,
      "page": "1",
      "nlp": {"event": nl_json['event'], "city": nl_json['city'][0], "country": nl_json['country'][0], "year": int(nl_json['date']['year']), "month": int(nl_json['date']['month']),
              "day": int(nl_json['date']['day']), "magnitude": nl_json['Number'], "comparative": nl_json['comparative'], "point": False, "latitude": "null", "lognitude": "null"}
    }

    return sf_json_format

  except:
    return "\n\n=== AN UNEXPECTED ERROR HAS OCCURED. PLEASE EXECUTE AGAIN THE SCRIPT OR COMMUNICATE WITH THE DEVELOPER TEAM === \n\n"



def main(sentence):
      
    nl_data = natural_language_module(sentence)
    nl_json = process_json_sf(nl_data, sentence)
    nl_json_with_null = replace_zero_with_null(nl_json)
    
    return nl_json_with_null