[authors-tools] Fix Orcid email error

services/authors-tools/Dockerfile

@@ -4,6 +4,7 @@ FROM python:3.9-slim-bullseye AS dvcfiles
 WORKDIR /dvc
 RUN apt-get update && apt-get -y --no-install-recommends install git=1:2.30.2-1+deb11u5 && apt-get clean && rm -rf /var/lib/apt/lists/*
 RUN pip install --no-cache-dir dvc[webdav]==3.42.0
+RUN pip install "pathspec==0.10.3"
 RUN --mount=type=secret,id=webdav_login \
     --mount=type=secret,id=webdav_password \
     --mount=type=secret,id=webdav_url \

services/authors-tools/v1/orcid-disambiguation/disambiguate.py

@@ -8,9 +8,13 @@
 from requests.auth import HTTPBasicAuth
 import os
 
+
 def getPoints(liste):
     return liste[2]
 
+def getSecondaryPoints(liste):
+    return liste[3]
+
 
 class disambiguate:
     def __init__(self, infoDic, nameDepth = 20, worksDepth = 20):
@@ -166,9 +170,13 @@ def splitFirstName(self,name):
         return choice
 
     def checkEmail(self,email):
+        mails = []
         for em in self.infoDic["email"]:
-            if em == email:
-                return True,em
+            for mail in email.split(","):
+                if em == mail:
+                    mails.append(em)
+        if len(mails) > 0:
+            return True,em
         return False,0
 
     def disambiguation(self):
@@ -183,7 +191,7 @@ def disambiguation(self):
             if "email" in self.infoDic: #check email
                 end,em = self.checkEmail(personInfos["email"])
                 if end:
-                    return [[orcid,["Email "+em],100]]
+                    return [[orcid,["Email "+em],100]][0][0]
 
             works = self.getWorksFromOrcid(orcid)
             time.sleep(self.timeBetweenrequest)
@@ -194,7 +202,7 @@ def disambiguation(self):
                     for tit in [title.lower() for title in worksInfo["title"]]:
                         ratio = SequenceMatcher(None, title.lower(), tit).ratio() #check similarity between title
                         if ratio > 0.7:
-                            return [[orcid,["title "+tit],100]]            
+                            return [[orcid,["title "+tit],100]][0][0]           
 
             if "coAuthors" in self.infoDic: #check coAuthors
                 authorsPutcode = []
@@ -206,7 +214,7 @@ def disambiguation(self):
                     choices = self.splitFirstName(author)
                     for choice in choices:
                         if choice in [auth.lower() for auth in authors]:
-                            return [[orcid,["Co-authors "+choice],100]]
+                            return [[orcid,["Co-authors "+choice],100]][0][0]
 
             if "affiliations" in self.infoDic: #check affiliations
                 for affiliation in self.infoDic["affiliations"]:
@@ -226,14 +234,25 @@ def disambiguation(self):
                     points += 5
                     matchArg.append("Match LastName ")                    
 
+            personInfos["secondaryScore"] = len(worksInfo["title"])
             personInfos["points"] = points
             personInfos["matchArg"] = matchArg    
 
         #check first and last name
         finalReturn = []
         for personInfos in personsInfos:
             if personInfos["points"] != 0:
-                finalReturn.append([personInfos["orcid"],personInfos["matchArg"],personInfos["points"]])
-
+                finalReturn.append([personInfos["orcid"],personInfos["matchArg"],
+                                    personInfos["points"], personInfos["secondaryScore"]])
+        #Better score first
         finalReturn.sort(key=getPoints,reverse=True)
-        return finalReturn
+        #if multiples sort based on secondary score
+        if len(finalReturn) == 0:
+            return "None"
+        max_score = finalReturn[0][2]
+        finalReturn = [t for t in finalReturn if t[2] == max_score]
+        if len(finalReturn) == 1:
+            return finalReturn[0][0]
+        else:
+            finalReturn.sort(key=getSecondaryPoints, reverse=True)
+            return [t[0] for t in finalReturn]

services/authors-tools/v1/orcid-disambiguation/orcidDisambiguation.py

@@ -17,11 +17,7 @@ def main(nameDepth = 20, worksDepth = 20):
         res = []
         for info in data['value'].copy():
             db = disambiguate(info,nameDepth=nameDepth, worksDepth=worksDepth )
-            result = db.disambiguation()
-            if len(result)>0:
-                res.append(result[0][0])
-            else:
-                res.append("None")
+            res.append(db.disambiguation())
         data["value"] = res
         sys.stdout.write(json.dumps(data))
         sys.stdout.write('\n')