fix(email_extraction): remove extra dots at the end of email addresses + extract some email addresses on two lines

- strip addresses of extra "."
- replace @\n with @ in fulltext before extraction to include cases like the following:
    you@
    email.com

Author: valearna

setup.py

@@ -5,7 +5,7 @@
 
 setuptools.setup(
     name="wbtools",
-    version="1.2.14",
+    version="1.2.15",
     author="Valerio Arnaboldi",
     author_email="valearna@caltech.edu",
     description="Interface to WormBase (www.wormbase.org) curation data, including literature management and NLP "

tests/lib/entity_extraction/test_email_addresses.py

@@ -1,6 +1,9 @@
 import unittest
+import os
 
+from tests.config_reader import read_db_config, read_tazendra_config
 from wbtools.lib.nlp.entity_extraction.email_addresses import get_email_addresses_from_text
+from wbtools.literature.corpus import CorpusManager
 
 
 class TestEmailAddresses(unittest.TestCase):
@@ -11,6 +14,24 @@ def test_get_email_addresses_from_text(self):
         addr = get_email_addresses_from_text(text)
         self.assertEqual(len(addr), 2)
 
+    @unittest.skipIf(not os.path.exists(os.path.join(os.path.dirname(os.path.abspath(__file__)), "../..", "data",
+                                                     "local_config", "db.cfg")), "Test DB config file not present")
+    def test_get_email_addresses_from_paper(self):
+        config = read_db_config()
+        tazendra_config = read_tazendra_config()
+        cm = CorpusManager()
+        cm.load_from_wb_database(db_name=config["wb_database"]["db_name"], db_user=config["wb_database"]["db_user"],
+                                 db_password=config["wb_database"]["db_password"],
+                                 db_host=config["wb_database"]["db_host"],
+                                 ssh_user=tazendra_config["ssh"]["ssh_user"],
+                                 ssh_passwd=tazendra_config["ssh"]["ssh_password"],
+                                 paper_ids=['00062455'])
+        email_addresses = get_email_addresses_from_text(cm.get_paper('00062455').get_text_docs(
+            include_supplemental=False, return_concatenated=True))
+        email_addresses_in_wb = cm.get_paper('00062455').get_authors_with_email_address_in_wb()
+        self.assertEqual(len(email_addresses), 3)
+        self.assertGreaterEqual(len(email_addresses_in_wb), 2)
+
 
 if __name__ == '__main__':
     unittest.main()

wbtools/lib/nlp/entity_extraction/email_addresses.py

@@ -4,11 +4,11 @@
 
 
 def get_email_addresses_from_text(text):
-    all_addresses = re.findall(EMAIL_ADDRESS_REGEX, text)
+    all_addresses = re.findall(EMAIL_ADDRESS_REGEX, text.replace('@\n', '@'))
     if not all_addresses:
         text = text.replace(". ", ".")
-        all_addresses = re.findall(EMAIL_ADDRESS_REGEX, text)
-        all_addresses = [address.strip(".") for address in all_addresses]
+        all_addresses = re.findall(EMAIL_ADDRESS_REGEX, text.replace('@\n', '@'))
+    all_addresses = [address.strip(".") for address in all_addresses]
     added_addresses = set()
     return_addresses = []
     for address in all_addresses: