config fix, added new skills as suggested on main repo, added linkedi…

…n link

pyresparser/resume_parser.py

@@ -62,6 +62,8 @@ def __get_basic_details(self):
                     self.__noun_chunks,
                     self.__skills_file
                 )
+
+        linkedin = utils.extract_linkedin(self.__text)
         # edu = utils.extract_education(
         #               [sent.string.strip() for sent in self.__nlp.sents]
         #       )
@@ -121,6 +123,8 @@ def __get_basic_details(self):
         self.__details['no_of_pages'] = utils.get_number_of_pages(
                                             self.__resume
                                         )
+
+        self.__details['linkedin'] = linkedin
         return
 
 

pyresparser/utils.py

@@ -494,3 +494,15 @@ def extract_experience(resume_text):
         if x and 'experience' in x.lower()
     ]
     return x
+
+def extract_linkedin(text):
+    '''
+    Helper function to extract linkedin from text
+    :param text: plain text extracted from resume file
+    '''
+    linkedin = re.findall(r'([\s]+linkedin.com[^\s]+)', text)
+    if linkedin:
+        try:
+            return linkedin[0][0]
+        except IndexError:
+            return None

setup.py

@@ -46,7 +46,7 @@
         'requests>=2.22.0',
         'six>=1.12.0',
         'sortedcontainers>=2.1.0',
-        'spacy>=2.1.4',
+        'spacy>=2.1.4,<=2.3.5',
         'srsly>=0.0.7',
         'thinc>=7.0.4',
         'tqdm>=4.32.2',

test_name.py

@@ -39,3 +39,8 @@ def test_local_name():
 def test_local_phone_number():
     data = get_local_data()
     assert '8087996634' == data['mobile_number']
+
+def test_linkedin_url():
+    data = get_remote_data()
+    # no valid linkedin url yet
+    assert None == data[0]['linkedin']