Provide __str__ where possible for different views

* #306 Fix of AbstractRetrieval-META

* #306 Fix of AbstractRetrieval-REF

* #306 Fix AuthorRetrieval - Light. The error in the __str__ was a result of the innability of many properties to retrieve the information of the LIGHT view. The reason is that the LIGHT view response has a different structure. Insead of changing the __str__ function some fixes to the properties were conducted.

* #306 Fix of AuthorRetrieval - Metrics. The __str__ was changed to return a string version conditional on the view.

* #306 Adjustment to tests and minor changes

* #306 AbstractRetrieval corrections

* #306 AuthorRetrieval requested changes

* #306 test AuthorRetrieval requested chnages

* #306 Parse content requested changes

* #306 Minor review changes

* #306 Minor requested changes II

Author: nils-herrmann

pybliometrics/scopus/abstract_retrieval.py

@@ -24,7 +24,7 @@ def affiliation(self) -> Optional[List[NamedTuple]]:
         aff = namedtuple('Affiliation', 'id name city country')
         affs = listify(self._json.get('affiliation', []))
         for item in affs:
-            new = aff(id=int(item['@id']), name=item.get('affilname'),
+            new = aff(id=make_int_if_possible(item.get('@id')), name=item.get('affilname'),
                       city=item.get('affiliation-city'),
                       country=item.get('affiliation-country'))
             out.append(new)
@@ -747,29 +747,55 @@ def __str__(self):
         Assumes the document is a journal article and was loaded with
         view="META_ABS" or view="FULL".
         """
-        date = self.get_cache_file_mdate().split()[0]
-        # Authors
-        if self.authors:
-            if len(self.authors) > 1:
-                authors = _list_authors(self.authors)
+        def convert_citedbycount(entry):
+                try:
+                    return float(entry.citedbycount) or 0
+                except (ValueError, TypeError):
+                    return 0
+            
+        def get_date(coverDate):
+            try:
+                return coverDate[:4]
+            except TypeError:
+                return None
+                
+        if self._view in ('FULL', 'META_ABS', 'META'):
+            date = self.get_cache_file_mdate().split()[0]
+            # Authors
+            if self.authors:
+                if len(self.authors) > 1:
+                    authors = _list_authors(self.authors)
+                else:
+                    a = self.authors[0]
+                    authors = str(a.given_name) + ' ' + str(a.surname)
             else:
-                a = self.authors[0]
-                authors = str(a.given_name) + ' ' + str(a.surname)
-        else:
-            authors = "(No author found)"
-        # All other information
-        s = f'{authors}: "{self.title}", {self.publicationName}, {self.volume}'
-        if self.issueIdentifier:
-            s += f'({self.issueIdentifier})'
-        s += ', '
-        s += _parse_pages(self)
-        s += f'({self.coverDate[:4]}).'
-        if self.doi:
-            s += f' https://doi.org/{self.doi}.\n'
-        s += f'{self.citedby_count} citation(s) as of {date}'
-        if self.affiliation:
-            s += "\n  Affiliation(s):\n   "
-            s += '\n   '.join([aff.name for aff in self.affiliation])
+                authors = "(No author found)"
+            # All other information
+            s = f'{authors}: "{self.title}", {self.publicationName}, {self.volume}'
+            if self.issueIdentifier:
+                s += f'({self.issueIdentifier})'
+            s += ', '
+            s += _parse_pages(self)
+            s += f'({self.coverDate[:4]}).'
+            if self.doi:
+                s += f' https://doi.org/{self.doi}.\n'
+            s += f'{self.citedby_count} citation(s) as of {date}'
+            if self.affiliation:
+                s += "\n  Affiliation(s):\n   "
+                s += '\n   '.join([aff.name for aff in self.affiliation])
+        
+        elif self._view in ('REF'):
+            # Sort reference list by citationcount
+            top_n = 5
+            references = sorted(self.references, key=convert_citedbycount, reverse=True)
+
+            top_references = [f'{reference.title} ({get_date(reference.coverDate)}). '+
+                              f'EID: {reference.id}' for reference in references[:top_n]]
+
+            s = f'A total of {self.refcount} references were found. '
+            s += f'Top {top_n} references:\n\t'
+            s += '\n\t'.join(top_references)
+
         return s
 
     def get_bibtex(self) -> str:

pybliometrics/scopus/author_retrieval.py

@@ -23,8 +23,13 @@ def affiliation_current(self) -> Optional[List[NamedTuple]]:
         when it lookes correct in the web view.  In this case please request
         a correction.
         """
-        affs = chained_get(self._profile, ["affiliation-current", "affiliation"])
-        return parse_affiliation(affs)
+        if self._view in ('STANDARD', 'ENHANCED'):
+            affs = chained_get(self._profile, ["affiliation-current", "affiliation"])
+        elif self._view == 'LIGHT':
+            affs = self._json.get('affiliation-current')
+        else:
+            return None
+        return parse_affiliation(affs, self._view)
 
     @property
     def affiliation_history(self) -> Optional[List[NamedTuple]]:
@@ -128,7 +133,21 @@ def identifier(self) -> int:
     @property
     def indexed_name(self) -> Optional[str]:
         """Author's name as indexed by Scopus."""
-        return html_unescape(chained_get(self._profile, ['preferred-name', 'indexed-name']))
+        if self._view in ('STANDARD', 'ENHANCED'):
+            indexed_name = html_unescape(chained_get(self._profile, ['preferred-name', 'indexed-name']))
+        elif self._view == 'LIGHT':
+            # Try to get indexed name from name-variants
+            name_variants = chained_get(self._json, ['name-variants', 'name-variant'])
+            if name_variants:
+                indexed_name = chained_get(name_variants[0], ['name-variant', 'indexed-name'])
+            else:
+                # In case of no name-variants get name from preferred-name
+                preferred_name = self._json.get('preferred-name')
+                indexed_name = ' '.join([preferred_name.get('initials', ''), preferred_name.get('surname', '')])
+        else:
+            indexed_name = None
+        
+        return indexed_name
 
     @property
     def initials(self) -> Optional[str]:
@@ -155,12 +174,23 @@ def orcid(self) -> Optional[str]:
 
     @property
     def publication_range(self) -> Optional[Tuple[int, int]]:
-        """Tuple containing years of first and last publication."""
-        r = self._profile.get('publication-range')
-        try:
-            return int(r['@start']), int(r['@end'])
-        except TypeError:
-            return None
+        """Tuple containing years of first and last publication."""        
+        if self._view in ('STANDARD', 'ENHANCED', 'LIGHT'):
+            if self._view in ('STANDARD', 'ENHANCED'):
+                r = self._profile.get('publication-range')
+                start = '@start'
+                end = '@end'
+            elif self._view == 'LIGHT':
+                r = self._json.get('publication-range')
+                start = 'start'
+                end = 'end'
+            
+            try:
+                return int(r.get(start)), int(r.get(end))
+            except TypeError:
+                return None
+            
+        return None
 
     @property
     def scopus_author_link(self) -> Optional[str]:
@@ -269,13 +299,19 @@ def __init__(self,
 
     def __str__(self):
         """Return a summary string."""
-        date = self.get_cache_file_mdate().split()[0]
-        main_aff = self.affiliation_current[0]
-        s = f"{self.indexed_name} from {main_aff.preferred_name} in "\
-            f"{main_aff.country},\npublished {int(self.document_count):,} "\
-            f"document(s) since {self.publication_range[0]} "\
-            f"\nwhich were cited by {int(self.cited_by_count):,} author(s) in "\
-            f"{int(self.citation_count):,} document(s) as of {date}"
+        if self._view in ('STANDARD', 'ENHANCED', 'LIGHT'):
+            date = self.get_cache_file_mdate().split()[0]
+            main_aff = self.affiliation_current[0]
+            s = f"{self.indexed_name} from {main_aff.preferred_name} in "\
+                f"{main_aff.country},\npublished {int(self.document_count):,} "\
+                f"document(s) since {self.publication_range[0]} "\
+                f"\nwhich were cited by {int(self.cited_by_count):,} author(s) in "\
+                f"{int(self.citation_count):,} document(s) as of {date}"
+        elif self._view == 'METRICS':
+            s = f'Author with ID {self._id}\n'\
+                f'published {int(self.document_count):,} document(s)\n'\
+                f'which were cited by {int(self.cited_by_count):,} author(s) '\
+                f'in {int(self.citation_count):,} document(s)'
         return s
 
     def get_coauthors(self) -> Optional[List[NamedTuple]]:

pybliometrics/scopus/tests/test_AuthorRetrieval.py

@@ -18,23 +18,34 @@
 
 def test_affiliation_current():
     assert_equal(metrics.affiliation_current, None)
-    assert_equal(light.affiliation_current, None)
+
     order = 'id parent type relationship afdispname preferred_name '\
             'parent_preferred_name country_code country address_part city '\
             'state postal_code org_domain org_URL'
-    aff = namedtuple('Affiliation', order)
-    expected = aff(id=110785688, parent=60027950, type='dept',
-        relationship='author', afdispname=None, country='United States',
-        preferred_name='Department of Chemical Engineering',
-        parent_preferred_name='Carnegie Mellon University', country_code='usa',
+    aff = namedtuple('Affiliation', order, defaults=(None,) * len(order.split()))
+
+    expected_std_enh = aff(id=60027950, parent=None, type='parent',
+        relationship='author', afdispname=None, preferred_name='Carnegie Mellon University',
+        parent_preferred_name=None, country_code='usa', country='United States',
         address_part='5000 Forbes Avenue', city='Pittsburgh', state='PA',
-        postal_code='15213-3890', org_domain='cmu.edu',
-        org_URL='https://www.cmu.edu/')
-    for a in (standard, enhanced):
+        postal_code='15213-3890', org_domain='cmu.edu', org_URL='https://www.cmu.edu/')
+
+    expected_lgh = aff(id=None, parent=None, type=None,
+        relationship=None, afdispname=None, preferred_name='Carnegie Mellon University',
+        parent_preferred_name=None, country_code=None, country='United States',
+        address_part=None, city='Pittsburgh', state=None,
+        postal_code=None, org_domain=None, org_URL=None)
+
+    for a in (standard, enhanced, light):
         received = a.affiliation_current
         assert_true(isinstance(received, list))
         assert_true(len(received) >= 1)
-        assert_true(expected in received)
+
+    for a in (standard, enhanced):
+        received = a.affiliation_current
+        assert_true(expected_std_enh in received)
+
+    assert_true(expected_lgh in light.affiliation_current)
 
 
 def test_affiliation_history():
@@ -199,7 +210,7 @@ def test_identifier():
 
 def test_indexed_name():
     assert_equal(metrics.indexed_name, None)
-    assert_equal(light.indexed_name, None)
+    assert_equal(light.indexed_name, 'Kitchin J.')
     assert_equal(standard.indexed_name, 'Kitchin J.')
     assert_equal(enhanced.indexed_name, 'Kitchin J.')
 
@@ -231,8 +242,7 @@ def test_orcid():
 
 def test_publication_range():
     assert_equal(metrics.publication_range, None)
-    assert_equal(light.publication_range, None)
-    for a in (standard, enhanced):
+    for a in (standard, enhanced, light):
         assert_equal(a.publication_range[0], 1995)
         assert_true(a.publication_range[1] >= 2021)
 

pybliometrics/scopus/utils/parse_content.py

@@ -140,32 +140,40 @@ def make_search_summary(self, keyword, results, joiner="\n    "):
     return s
 
 
-def parse_affiliation(affs):
+def parse_affiliation(affs, view):
     """Helper function to parse list of affiliation-related information."""
     order = 'id parent type relationship afdispname preferred_name '\
             'parent_preferred_name country_code country address_part city '\
             'state postal_code org_domain org_URL'
-    aff = namedtuple('Affiliation', order)
+    aff = namedtuple('Affiliation', order, defaults=(None,) * len(order.split()))
     out = []
-    for item in listify(affs):
-        if not item:
-            continue
-        doc = item.get('ip-doc', {}) or {}
-        address = doc.get('address', {}) or {}
-        try:
-            parent = int(item['@parent'])
-        except KeyError:
-            parent = None
-        new = aff(id=int(item['@affiliation-id']), parent=parent,
-            type=doc.get('@type'), relationship=doc.get('@relationship'),
-            afdispname=doc.get('@afdispname'),
-            preferred_name=doc.get('preferred-name', {}).get('$'),
-            parent_preferred_name=doc.get('parent-preferred-name', {}).get('$'),
-            country_code=address.get('@country'), country=address.get('country'),
-            address_part=address.get("address-part"), city=address.get('city'),
-            postal_code=address.get('postal-code'), state=address.get('state'),
-            org_domain=doc.get('org-domain'), org_URL=doc.get('org-URL'))
+
+    if view in ('STANDARD', 'ENHANCED'):
+        for item in listify(affs):
+            if not item:
+                continue
+            doc = item.get('ip-doc', {}) or {}
+            address = doc.get('address', {}) or {}
+            try:
+                parent = int(item['@parent'])
+            except KeyError:
+                parent = None
+            new = aff(id=int(item['@affiliation-id']), parent=parent,
+                type=doc.get('@type'), relationship=doc.get('@relationship'),
+                afdispname=doc.get('@afdispname'),
+                preferred_name=doc.get('preferred-name', {}).get('$'),
+                parent_preferred_name=doc.get('parent-preferred-name', {}).get('$'),
+                country_code=address.get('@country'), country=address.get('country'),
+                address_part=address.get("address-part"), city=address.get('city'),
+                postal_code=address.get('postal-code'), state=address.get('state'),
+                org_domain=doc.get('org-domain'), org_URL=doc.get('org-URL'))
+            out.append(new)
+    elif view == 'LIGHT':
+        new = aff(preferred_name=affs.get('affiliation-name'),
+                  city=affs.get('affiliation-city'),
+                  country=affs.get('affiliation-country'))
         out.append(new)
+
     return out or None