Update REcordHepPubli to fix an issue in process_publication_info

modules/store_tools/recordheppubli.py

 """ store_tools.recordheppubli
 
 """
+import logging
+
 from filters import CLEAN_COLLABORATION
 from pandas import DataFrame
 from .recordhep import RecordHep
@@ -51,6 +53,8 @@ class RecordHepPubli(RecordHep, PluginAuthors, PluginPublicationInfo):
 
         super().__init__(recjson)
 
+        self.logger = logging.getLogger("web2py.app.limbra")
+
         self._last_fmt_author = "Last, First"
         self._process_authors()
         self._process_publication_info()
@@ -83,6 +87,8 @@ class RecordHepPubli(RecordHep, PluginAuthors, PluginPublicationInfo):
             is not defined.
 
         """
+        self.logger.debug("    process authors")
+
         authors = self.get("authors", None)
 
         if authors is None:
@@ -159,20 +165,26 @@ class RecordHepPubli(RecordHep, PluginAuthors, PluginPublicationInfo):
               and volume.
 
         """
+        self.logger.debug("    process publication info")
+
+        cols = ["title",
+                "volume",
+                "year",
+                "pagination"]
+
         data = self.get("publication_info", None)
 
         if data is None:
-            cols = ["title",
-                    "volume",
-                    "year",
-                    "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]
 
+        if len(data) == 0:
+            self.df_info = DataFrame([[""] * len(cols)], columns=cols)
+            return
+
         # convert data to DataFrame with a well know structure
         df = (DataFrame(data)
               .astype({"year": str})