emails

api/server.py

@@ -8,6 +8,7 @@
 from waitress import serve
 
 from names_dataset import NameDataset, NameWrapper
+from names_dataset.emails import extract_names_from_email
 
 logger = logging.getLogger(__name__)
 logging.basicConfig(
@@ -46,6 +47,37 @@ def str2bool(s: Union[bool, str]) -> bool:
         return False
 
 
+@app.route('/split', methods=['GET'])
+def split():
+    try:
+        req = request
+        q = req.args.get('q')
+        if q is None:
+            return generate_output(
+                'provide a parameter q, for example '
+                '[email protected] or philipperemy', status=False
+            )
+        else:
+            first_name, last_name = extract_names_from_email(nd, q)
+            if first_name is not None:
+                result_first_name = nd.search(first_name)['first_name']
+            else:
+                result_first_name = None
+            if last_name is not None:
+                result_last_name = nd.search(last_name)['last_name']
+            else:
+                result_last_name = None
+            result_first_name['name'] = first_name
+            result_last_name['name'] = last_name
+            result = {
+                'first_name': result_first_name,
+                'last_name': result_last_name
+            }
+            return generate_output({'result': result}, status=True)
+    except Exception as e:
+        return generate_output({'error': str(e)}, status=True)
+
+
 @app.route('/country_codes', methods=['GET'])
 def country_codes():
     try:

names_dataset/emails.py

@@ -5,8 +5,6 @@
 
 from names_dataset import NameDataset
 
-nd = NameDataset()
-
 
 def _compute_score(ranks: Dict):
     values = {a: b for a, b in ranks['rank'].items() if b is not None}.values()
@@ -15,7 +13,7 @@ def _compute_score(ranks: Dict):
     return -min(values)
 
 
-def _score(candidate: str):
+def _score(nd: NameDataset, candidate: str):
     if len(candidate) == 0:
         return float('-inf')
     first_name = nd.search(candidate)['first_name']
@@ -33,8 +31,8 @@ def _score(candidate: str):
 
 
 # Function to infer the best split between first and last name
-def _infer_best_split(full_name: str):
-    max_score = _score(full_name)
+def _infer_best_split(nd: NameDataset, full_name: str):
+    max_score = _score(nd, full_name)
     best_split = (full_name, None)
 
     # Try all possible ways to split the full_name
@@ -43,7 +41,7 @@ def _infer_best_split(full_name: str):
         last = full_name[i:]
 
         # Calculate total score for the split
-        total_score = _score(first) + _score(last)
+        total_score = _score(nd, first) + _score(nd, last)
 
         # If this split has a higher score, update the best split
         if total_score > max_score:
@@ -53,14 +51,17 @@ def _infer_best_split(full_name: str):
     return best_split, max_score
 
 
-def _general_score(candidate: str):
+def _general_score(nd: NameDataset, candidate: str):
     c = nd.search(candidate)
     s1 = _compute_score(c['first_name'])
     s2 = _compute_score(c['last_name'])
     return max(s1, s2)
 
 
-def extract_names_from_email(email: str):
+def extract_names_from_email(nd: NameDataset, email: str):
+    if '@' not in email:
+        email += '@gmail.com'
+
     email = ''.join([e for e in list(email) if not e.isnumeric()])
 
     prefix, suffix = email.split('@')
@@ -76,7 +77,7 @@ def extract_names_from_email(email: str):
     for e in ['.', '_', '-']:
         if prefix.count(e) >= 2:
             c_list = prefix.split(e)
-            scores = [_general_score(c) for c in c_list]
+            scores = [_general_score(nd, c) for c in c_list]
             a, b = np.array(c_list)[np.argsort(scores)][-2:]
             email = f'{a}.{b}@{suffix}'
 
@@ -102,7 +103,7 @@ def extract_names_from_email(email: str):
 
     if not had_matched:
         prefix = email.split('@')[0]
-        (first_name, last_name), max_score = _infer_best_split(prefix)
+        (first_name, last_name), max_score = _infer_best_split(nd, prefix)
 
     if first_name is not None and len(first_name) == 1:
         first_name = None

tests/test_from_emails.py

@@ -1,5 +1,6 @@
 import unittest
 
+from names_dataset import NameDataset
 from names_dataset.emails import extract_names_from_email
 
 
@@ -28,6 +29,9 @@ def test_1(self):
             '[email protected]',
             '[email protected]',
         ]
+        inputs2 = []
+        for i in inputs:
+            inputs2.append(i.split('@')[0])
 
         outputs = [
             [None, None],
@@ -52,8 +56,9 @@ def test_1(self):
             ['philippe', 'remy'],
         ]
 
-        for input_, output_ in zip(inputs, outputs):
-            first_name, last_name = extract_names_from_email(input_)
+        nd = NameDataset()
+        for input_, output_ in zip(inputs2, outputs):
+            first_name, last_name = extract_names_from_email(nd, input_)
             print(input_)
             self.assertEqual(output_[0], first_name)
             self.assertEqual(output_[1], last_name)