Merge pull request #17 from materials-data-facility/mdfcc-dev

0.3.2

Author: jgaff

README.md

@@ -1,4 +1,6 @@
 # MDF Connect Client
+[![PyPI](https://img.shields.io/pypi/v/mdf_connect_client.svg)](https://pypi.python.org/pypi/mdf-connect-client) [![Build Status](https://travis-ci.org/materials-data-facility/connect_client.svg?branch=master)](https://travis-ci.org/materials-data-facility/connect_client) [![Coverage Status](https://coveralls.io/repos/github/materials-data-facility/connect_client/badge.svg?branch=master)](https://coveralls.io/github/materials-data-facility/connect_client?branch=master) [![Read the Docs](https://readthedocs.org/projects/mdf-connect-client/badge/?version=master)](http://mdf-connect-client.readthedocs.io/en/master/)
+
 The MDF Connect Client is the Python client to easily submit datasets to MDF Connect.
 
 # Installation
@@ -19,7 +21,7 @@ Documentation is available on [Read the Docs](https://mdf-connect-client.readthe
 
 
 # Requirements
-* The Connect Client requires Python 3.4 or greater.
+* The Connect Client requires Python 3.5 or greater.
 * To submit data to MDF Connect, you must have an account recognized by Globus Auth (including Google, ORCiD, many academic institutions, or a [free Globus ID](https://www.globusid.org/create)), and be a member of the [MDF Connect Convert Globus Group](https://app.globus.org/groups/cc192dca-3751-11e8-90c1-0a7c735d220a).
 
 

mdf_connect_client/mdfcc.py

@@ -3,6 +3,7 @@
 
 import globus_sdk
 import mdf_toolbox
+from nameparser import HumanName
 import requests
 
 
@@ -106,13 +107,8 @@ def create_dc_block(self, title, authors,
 
         Arguments:
             title (str or list of str): The title(s) of the dataset.
-            authors (str or list of str): The author(s) of the dataset. Format must be one of:
-
-                    * ``"Givenname Familyname"``
-                    * ``"Familyname, Givenname"``
-                    * ``"Familyname; Givenname"``
-
-                    No additional commas or semicolons are permitted.
+            authors (str or list of str): The author(s) of the dataset.
+                    The name will be automatically parsed into given name and family name.
             publisher (str): The publisher of the dataset (not an associated paper).
                     **Default:** The Materials Data Facility.
             publication_year (int or str): The year of dataset publication.
@@ -174,25 +170,16 @@ def create_dc_block(self, title, authors,
             affiliations = [affiliations] * len(authors)
         creators = []
         for auth, affs in zip(authors, affiliations):
-            if auth.find(",") >= 0:
-                family, given = auth.split(",", 1)
-            elif auth.find(";") >= 0:
-                family, given = auth.split(";", 1)
-            elif auth.find(" ") >= 0:
-                given, family = auth.split(" ", 1)
-            else:
-                given = auth
-                family = ""
-            if not isinstance(affs, list):
-                affs = [affs]
-
-            family = family.strip()
-            given = given.strip()
+            name = HumanName(auth)
+            given = "{} {}".format(name.first, name.middle).strip()
+            family = "{} {}".format(name.last, name.suffix).strip()
             creator = {
-                "creatorName": "{}, {}".format(family, given),
+                "creatorName": "{}, {}".format(family, given).strip(" ,"),
                 "familyName": family,
                 "givenName": given
             }
+            if not isinstance(affs, list):
+                affs = [affs]
             if affs:
                 creator["affiliations"] = affs
             creators.append(creator)

setup.py

@@ -2,13 +2,14 @@
 
 setup(
     name='mdf_connect_client',
-    version='0.3.1',
+    version='0.3.2',
     packages=['mdf_connect_client'],
     description='Materials Data Facility Connect Client',
     long_description=("The MDF Connect Client is the Python client to easily submit"
                       " datasets to MDF Connect."),
     install_requires=[
         "mdf-toolbox>=0.2.6",
+        "nameparser>=1.0.4",
         "requests>=2.18.4"
     ],
     python_requires=">=3.5",

tests/test_connect_client.py

@@ -22,7 +22,7 @@ def test_create_dc_block():
     # Full test, no lists
     mdf.create_dc_block(
         title="Connect Title",
-        authors="Data Facility; Materials",
+        authors="Data Facility, Materials",
         affiliations="UChicago",
         publisher="Globus",
         publication_year="2017",
@@ -70,7 +70,7 @@ def test_create_dc_block():
     # Full test, all lists
     mdf.create_dc_block(
         title=["Connect Title", "Other Title"],
-        authors=["Data Facility; Materials", "Blaiszik, Ben", "Jonathon Gaff"],
+        authors=["Data Facility, Materials", "Blaiszik, Ben", "Jonathon Gaff"],
         affiliations=["UChicago", "Argonne"],
         publisher="Globus",
         publication_year="2017",
@@ -143,18 +143,18 @@ def test_create_dc_block():
     # Minimum test
     mdf.create_dc_block(
         title="Project One",
-        authors=["Senior Programmer", "Programmer, Junior"]
+        authors=["Artemis Moonshot", "Landing, Apollo"]
     )
     assert mdf.dc == {
         'creators': [{
-                'creatorName': 'Programmer, Senior',
-                'familyName': 'Programmer',
-                'givenName': 'Senior'
+                'creatorName': 'Moonshot, Artemis',
+                'familyName': 'Moonshot',
+                'givenName': 'Artemis'
             },
             {
-                'creatorName': 'Programmer, Junior',
-                'familyName': 'Programmer',
-                'givenName': 'Junior'
+                'creatorName': 'Landing, Apollo',
+                'familyName': 'Landing',
+                'givenName': 'Apollo'
             }
         ],
         'publicationYear': '2019',