Handle PubmedData being None in pubmed client

indra/literature/pubmed_client.py

@@ -353,6 +353,15 @@ def get_full_xml_by_pmids(
     ------
     RuntimeError
         If the edirect CLI utilities are not installed or not found on PATH.
+
+    Notes
+    -----
+    - This function requires the edirect command line utilities to be installed
+      and visible on your PATH. See https://www.ncbi.nlm.nih.gov/books/NBK179288/
+      for instructions.
+    - Note that the output is sorted by PMID numerically e.g.,
+      10, 11, 20, 22, 1000 (and not lexicographically e.g., 10, 1000, 11, 20, 22)
+      without regard to the order in which the pmids are passed in.
     """
     # Have to use lxml.etree because the XML returned by efetch is not properly
     # formatted for ET.XML
@@ -371,11 +380,10 @@ def get_full_xml_by_pmids(
                            "for instructions.")
 
     tree = lxml_etree.fromstring(xml_bytes, parser=parser)
+    if tree is None:
+        raise RuntimeError("Could not parse XML returned by efetch.")
     # Each article is in a <PubmedArticle> tag, encapsulated in a
     # <PubmedArticleSet> tag.
-    # Note that the <PubmedArticle> tags are sorted by PMID numerically e.g.,
-    # 10, 11, 20, 1000, and not lexicographically e.g., 10, 1000, 11, 20,
-    # regardless of the order in which the pmids are passed
     if fname is not None:
         pretty_save_xml(tree, fname)
     return tree
@@ -426,6 +434,9 @@ def get_abstract(pubmed_id, prepend_title=True):
 
 # A function to get the text for the element, or None if not found
 def _find_elem_text(root, xpath_string):
+    if root is None:
+        logger.warning("Root is None when trying to find element with xpath: %s" % xpath_string)
+        return None
     elem = root.find(xpath_string)
     return None if elem is None else elem.text
 
@@ -673,7 +684,7 @@ def _get_references(reference_list, only_pmid=True):
     return references
 
 
-def _get_article_info(medline_citation, pubmed_data, detailed_authors=False):
+def _get_article_info(medline_citation, pubmed_data=None, detailed_authors=False):
     article = medline_citation.find('Article')
     pmid = _find_elem_text(medline_citation, './PMID')
     pii = _find_elem_text(article,
@@ -684,11 +695,12 @@ def _get_article_info(medline_citation, pubmed_data, detailed_authors=False):
                           './ELocationID[@EIdType="doi"][@ValidYN="Y"]')
 
     # ...and if that doesn't work, look in the ArticleIdList
-    if doi is None:
+    if doi is None and pubmed_data is not None:
         doi = _find_elem_text(pubmed_data, './/ArticleId[@IdType="doi"]')
 
     # Try to get the PMCID
-    pmcid = _find_elem_text(pubmed_data, './/ArticleId[@IdType="pmc"]')
+    if pubmed_data is not None:
+        pmcid = _find_elem_text(pubmed_data, './/ArticleId[@IdType="pmc"]')
 
     # Title
     title = _get_title_from_article_element(article)
@@ -763,7 +775,11 @@ def get_metadata_from_pubmed_article(
 
     Returns
     -------
-
+    : Dict
+        A dict containing the following fields: 'doi', 'title', 'authors',
+        'journal_title', 'journal_abbrev', 'journal_nlm_id', 'issn_list',
+        'page', 'volume', 'issue', 'issue_pub_date', 'mesh_annotations',
+        'publication_date', 'abstract', 'publication_types' and 'references'.
     """
     medline_citation = pubmed_article.find('./MedlineCitation')
     pubmed_data = pubmed_article.find('PubmedData')
@@ -777,13 +793,14 @@ def get_metadata_from_pubmed_article(
     if mesh_annotations:
         context_info = _get_annotations(medline_citation)
         result.update(context_info)
-    if references_included:
+    if references_included and pubmed_data is not None:
         references = _get_references(pubmed_data.find('ReferenceList'),
                                      only_pmid=(references_included == 'pmid'))
         result['references'] = references
 
-    publication_date = _get_pubmed_publication_date(pubmed_data)
-    result['publication_date'] = publication_date
+    if pubmed_data is not None:
+        publication_date = _get_pubmed_publication_date(pubmed_data)
+        result['publication_date'] = publication_date
 
     # Get the abstracts if requested
     if get_abstracts:

indra/preassembler/grounding_mapper/gilda.py

@@ -15,7 +15,7 @@
 logger = logging.getLogger(__name__)
 
 grounding_service_url = get_config('GILDA_URL', failure_ok=True) \
-    if has_config('GILDA_URL') else 'http://grounding.indra.bio/'
+    if has_config('GILDA_URL') else 'https://grounding.indra.bio/'
 
 
 def get_grounding(