Improve RecordHepPubli._process_publication_info

modules/store_tools/recordhep.py

 """recordhep.py
 
 """
+import logging
 import pprint
 
 
@@ -20,6 +21,8 @@ class RecordHep(dict):
 
         super().__init__(recjson)
 
+        self.logger = logging.getLogger("web2py.app.limbra")
+
         # meta data
         # the authors of my institutes signing the record
         # string containing a list of name separated by a comma

modules/store_tools/recordheppubli.py

@@ -8,6 +8,37 @@ from store_tools.pluginauthors import PluginAuthors
 from store_tools.pluginpublicationinfo import PluginPublicationInfo
 
 
+def pages(row):
+    """Help function to build the pages argument
+
+    Args:
+        row (pandas.Series):
+            * artid (str)
+            * page_start (int)
+            * page_end (int)
+
+    Return
+        str:
+            * either 23 or 23-45
+            * empty string when information is missing
+
+    """
+    artid = row.get("artid", None)
+    pstart = row.get("page_start", None)
+    pend = row.get("page_end", None)
+
+    if pstart is None and pend is None and artid is None:
+        return ""
+
+    elif pstart is None and pend is None:
+        return artid
+
+    elif pend is None:
+        return f"{pstart}"
+
+    return f"{pstart}-{pend}"
+
+
 class RecordHepPubli(RecordHep, PluginAuthors, PluginPublicationInfo):
     """Article, preprint and proceeding from inspirehep.net version 2.
 
@@ -137,18 +168,23 @@ class RecordHepPubli(RecordHep, PluginAuthors, PluginPublicationInfo):
                     "pagination"]
 
             self.df_info = DataFrame([[""] * len(cols)], columns=cols)
-
             return
 
+        # filter data to keep only row with year information
+        data = [dct for dct in data if "year" in dct]
+
+        # convert data to DataFrame with a well know structure
         df = (DataFrame(data)
               .astype({"year": str})
-              .rename(columns={"artid": "pagination",
-                               "journal_title": "title",
-                               "journal_volume": "volume"}))
+              .rename(columns={"journal_title": "title",
+                               "journal_volume": "volume"},
+                      errors="ignore"))
 
-        columns = df.columns
+        # construction pagination columns
+        df["pagination"] = df.apply(pages, axis="columns")
 
         # erratum -- sort by year and volume
+        columns = df.columns
         if set(["year", "volume"]).issubset(columns):
             df = df.sort_values(["year", "volume"])