def __get_basic_details(self):
name = utils.extract_name(self.__resume)
email = utils.extract_email(self.__text)
mobile = utils.extract_mobile_number(self.__text)
skills = utils.extract_skills(self.__nlp, self.__noun_chunks)
edu = utils.extract_education(
[sent.string.strip() for sent in self.__nlp.sents])
experience = utils.extract_experience(self.__text)
entities = utils.extract_entity_sections(self.__text_raw)
self.__details['name'] = name
self.__details['email'] = email
self.__details['mobile_number'] = mobile
self.__details['skills'] = skills
# self.__details['education'] = entities['education']
self.__details['education'] = edu
self.__details['experience'] = experience
try:
self.__details['competencies'] = utils.extract_competencies(
self.__text_raw, entities['experience'])
self.__details[
'measurable_results'] = utils.extract_measurable_results(
self.__text_raw, entities['experience'])
except KeyError:
self.__details['competencies'] = []
self.__details['measurable_results'] = []
return
def get_details(self, resume_name):
resume_path = "Resumes/" + resume_name
ext = "pdf"
nlp = spacy.load('en_core_web_sm')
matcher = Matcher(nlp.vocab)
text_raw = utils.extract_text(resume_path, '.' + ext)
text = ' '.join(text_raw.split())
array = text.split()
topics = []
field_list = [
'OVERVIEW / CAREER OBJECTIVE / SUMMARY', 'KEY EXPERTISE / SKILLS',
'EDUCATION', 'AWARDS AND SCHOLARSHIPS', 'INTERNSHIPS', 'PROJECTS',
'ACHIEVEMENTS', 'SEMINARS / TRAININGS / WORKSHOPS',
'CO-CURRICULAR ACTIVITIES', 'EXTRA CURRICULAR ACTIVITIES',
'PERSONAL INTERESTS / HOBBIES', 'WEB LINKS', 'PERSONAL DETAILS'
]
for word in field_list:
if (text.find(word) >= 0):
topics.append(word)
content = {}
total_topics = len(topics)
for i in range(total_topics - 1):
string_to_find = topics[i] + '(.*)' + topics[i + 1]
result = re.search(string_to_find, text)
content[topics[i]] = result.group(1)
temp = topics[total_topics - 1] + '(.*)'
temp_res = re.search(temp, text)
content[topics[total_topics - 1]] = temp_res.group(1)
__full_text = nlp(text)
actual_marks = "CGPA: " + '(.*)' + "/ 10.00"
cgpa = re.search(actual_marks, content['EDUCATION'])
# DOMAIN RANKING
rank_text = content['KEY EXPERTISE / SKILLS'] + content['PROJECTS']
project_text = ResumeParse().clean_project(rank_text)
file_name = "rank/" + resume_name.split('.')[0] + ".txt"
f = open(file_name, "w+")
f.write(project_text)
f.close()
#FOR SKILLS
skills = ResumeParse().get_skills(content['KEY EXPERTISE / SKILLS'])
# name=utils.extract_name(__full_text,matcher)
email = utils.extract_email(text)
mobile = utils.extract_mobile_number(text)
details = {}
# details['name']=name
details['email'] = email
details['mobile'] = mobile
details['skills'] = skills
details['cgpa'] = cgpa.group(1)
return details, content
def __get_basic_details(self):
name = utils.extract_name(self.__nlp, matcher=self.__matcher)
email = utils.extract_email(self.__text)
mobile = utils.extract_mobile_number(self.__text)
skills = utils.extract_skills(self.__nlp, self.__noun_chunks)
edu = utils.extract_education([sent.string.strip() for sent in self.__nlp.sents])
entities = utils.extract_entity_sections_grad(self.__text_raw)
self.__details['name'] = name
self.__details['email'] = email
self.__details['mobile_number'] = mobile
self.__details['skills'] = skills
self.__details['education'] = edu
try:
self.__details['experience'] = entities['experience']
except KeyError:
self.__details['total_experience'] = 0
self.__details['no_of_pages'] = utils.get_number_of_pages(self.__resume)
return
def __get_basic_details(self):
name = utils.extract_name(self.__nlp, matcher=self.__matcher)
full_name = utils.get_first_name(name, self.__nlp)
gender = utils.get_gender(self.__nlp)
maritial_status = utils.get_maritial_status(self.__nlp)
passport_number = utils.get_passport_number(self.__text_raw)
date_of_birth = utils.extract_date_of_birth(self.__nlp, self.__text)
email = utils.extract_email(self.__text)
mobile = utils.extract_mobile_number(self.__text)
skills = utils.extract_skills(self.__nlp, self.__noun_chunks)
nationality = utils.get_nationality(self.__nlp)
languages = utils.extract_language(self.__nlp, self.__noun_chunks,
self.__languages_file)
num_of_companies = utils.extract_no_of_companies_worked_for(
self.__nlp, self.__noun_chunks, self.__companies_file)
hobbies = utils.extract_hobbies(self.__nlp, self.__noun_chunks,
self.__hobbies_file)
edu = utils.extract_education(
[sent.string.strip() for sent in self.__nlp.sents], self.__nlp,
self.__resume, date_of_birth)
entities = utils.extract_entity_sections_grad(self.__text_raw)
address = utils.extract_address(self.__nlp, self.__noun_chunks)
states = utils.extract_state(self.__nlp, self.__noun_chunks)
pincodes = utils.extract_pin(self.__nlp, self.__noun_chunks)
cities = utils.extract_cities(self.__nlp, self.__noun_chunks)
experience = utils.extract_experience_exceptional(
self.__nlp, self.__noun_chunks)
self.__details['name'] = name
self.__details['full_name'] = full_name
self.__details['gender'] = gender
self.__details['maritial_status'] = maritial_status
self.__details['passport_number'] = passport_number
self.__details['date_of_birth'] = date_of_birth
self.__details['email'] = email
self.__details['mobile_number'] = mobile
self.__details['skills'] = skills
self.__details['nationality'] = nationality
self.__details['languages'] = languages
self.__details['No. of companies'] = num_of_companies
self.__details['hobbies'] = hobbies
self.__details['education'] = edu
self.__details['address'] = address
self.__details['state'] = states
self.__details['pin'] = pincodes
self.__details['city'] = cities
self.__details['experience'] = experience
try:
#self.__details['experience'] = entities['experience']
try:
self.__details['competencies'] = 'none'
utils.extract_competencies(self.__text_raw,
entities['experience'])
self.__details['measurable_results'] = 'none'
utils.extract_measurable_results(self.__text_raw,
entities['experience'])
self.__details['total_experience'] = round(
utils.get_total_experience(entities['experience']) / 12, 2)
except KeyError:
self.__details['competencies'] = {}
self.__details['measurable_results'] = {}
self.__details['total_experience'] = 0
except KeyError:
self.__details['competencies'] = {}
self.__details['measurable_results'] = {}
self.__details['total_experience'] = 0
self.__details['no_of_pages'] = {
} #utils.get_number_of_pages(self.__resume)
#comented by vishal sharma 11/8/2019
'''
if len( self.__details['city'])>1:
cities = utils.extract_cities_exceptional(self.__nlp, self.__noun_chunks,self.__details['city'])
self.__details['city'] = cities
'''
if len(self.__details['pin']) > 1:
pincodes = utils.extract_pin_exceptional(self.__nlp,
self.__noun_chunks,
self.__details['pin'])
self.__details['pin'] = pincodes
return
import json
if not os.path.exists('./out'):
os.makedirs('./out')
fname = sys.argv[1]
utils.resume_file_name = fname.split('/')[-1]
print('Reading: %s' % fname)
print('Output will be written to: %s' %
('./out/' + utils.resume_file_name + '.json'))
content = utils.fread(fname, clean=True)
raw_content = utils.fread(fname, clean=False)
phone_numbers = utils.extract_phone_number(content)
emails = utils.extract_email(content)
names = utils.extract_names(content)
education = utils.get_education(raw_content)
skills = utils.get_skills(raw_content)
res = {
'name': names,
'email': emails,
'phone': phone_numbers,
'education': education,
'skills': skills
}
def __get_basic_details(self):
custom_entities = utils.extract_entities_wih_custom_model(
self.__custom_nlp)
name = utils.extract_name(self.__nlp, matcher=self.__matcher)
email = utils.extract_email(self.__text)
mobile = utils.extract_mobile_number(self.__text)
skills = utils.extract_skills(self.__nlp, self.__noun_chunks,
self.__skills_file)
edu = utils.extract_education(
[sent.string.strip() for sent in self.__nlp.sents])
entities = utils.extract_entity_sections_grad(self.__text_raw)
print(name)
# extract name
try:
self.__details['name'] = custom_entities['Name'][0]
except (IndexError, KeyError):
self.__details['name'] = name
#extract email
self.__details['email'] = email
# extract mobile number
self.__details['mobile_number'] = mobile
# extract skills
self.__details['skills'] = skills
# extract college name
try:
self.__details['college_name'] = entities['College Name']
except KeyError:
pass
# extract education Degree
try:
self.__details['degree'] = custom_entities['Degree']
except KeyError:
pass
# extract designation
try:
self.__details['designation'] = custom_entities['Designation']
except KeyError:
pass
# extract company names
try:
self.__details['company_names'] = custom_entities[
'Companies worked at']
except KeyError:
pass
try:
self.__details['experience'] = entities['experience']
try:
self.__details['total_experience'] = round(
utils.get_total_experience(entities['experience']) / 12, 2)
except KeyError:
self.__details['total_experience'] = 0
except KeyError:
self.__details['total_experience'] = 0
self.__details['no_of_pages'] = utils.get_number_of_pages(
self.__resume)
return