Handle empty search

CHANGELOG.md

@@ -1,5 +1,15 @@
 # Changelog
 
+## 1.1.1
+
+### Added
+
+- Test with empty papers list in `tests/test_telegrm.py`
+
+### Changed
+
+- Improve handling of papers list when it is empty
+
 ## 1.1.0
 
 ### Added

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "PaperBee"
-version = "1.1.0"
+version = "1.1.1"
 description = ""
 authors = ["Daniele Lucarelli <daniele.lucarelli@tum.de>", "Vladimir Shitov <vladimir.shitov@helmholtz-munich.de>"]
 repository = "https://github.com/theislab/paperbee"

src/PaperBee/papers/utils.py

@@ -43,44 +43,47 @@ def process_articles(self) -> None:
     def filter_columns(self) -> None:
         """Filters the DataFrame to include specific columns."""
         columns = ["databases", "publication_date", "title", "keywords", "url"]
-        self.articles = self.articles.loc[:, columns]
+
+        if self.articles.empty:
+            self.articles = pd.DataFrame(columns=columns)
+        else:
+            self.articles = self.articles.loc[:, columns]
 
     def extract_doi(self) -> None:
         """Extracts DOIs from URLs and adds them as a new column."""
-        self.articles["DOI"] = self.articles["url"].apply(lambda x: x[x.find("10.") :])
+        if not self.articles.empty:
+            self.articles["DOI"] = self.articles["url"].apply(lambda x: x[x.find("10.") :])
 
     def set_dates(self) -> None:
         """Sets the publication date and the date of processing."""
-        self.articles["Date"] = self.today_str
-        self.articles["PostedDate"] = self.articles["publication_date"]
+        if not self.articles.empty:
+            self.articles["Date"] = self.today_str
+            self.articles["PostedDate"] = self.articles["publication_date"]
 
     def determine_preprint_status(self) -> None:
         """Determines whether each article is a preprint based on its database."""
-        self.articles["IsPreprint"] = self.articles["databases"].apply(
-            lambda dbs: "FALSE" if "PubMed" in dbs else "TRUE"
-        )
+        if not self.articles.empty:
+            self.articles["IsPreprint"] = self.articles["databases"].apply(
+                lambda dbs: "FALSE" if "PubMed" in dbs else "TRUE"
+            )
 
     def rename_and_process_columns(self) -> None:
         """Renames columns and processes keywords."""
-        self.articles["Title"] = self.articles["title"]
-        self.articles["Keywords"] = self.articles["keywords"].apply(lambda kws: ", ".join(kw[2:] for kw in kws))
-        self.articles["URL"] = self.articles["url"]
+        if not self.articles.empty:
+            self.articles["Title"] = self.articles["title"]
+            self.articles["Keywords"] = self.articles["keywords"].apply(lambda kws: ", ".join(kw[2:] for kw in kws))
+            self.articles["URL"] = self.articles["url"]
 
     def select_last_columns(self) -> None:
         """Selects and rearranges the final set of columns for the DataFrame."""
-        self.articles["Preprint"] = None  # TODO add search for preprint of published articles
-        self.articles = self.articles[
-            [
-                "DOI",
-                "Date",
-                "PostedDate",
-                "IsPreprint",
-                "Title",
-                "Keywords",
-                "Preprint",
-                "URL",
-            ]
-        ]
+        expected_columns = ["DOI", "Date", "PostedDate", "IsPreprint", "Title", "Keywords", "Preprint", "URL"]
+        if self.articles.empty:
+            self.articles["Preprint"] = []
+            # Create empty DataFrame with expected columns
+            self.articles = pd.DataFrame(columns=expected_columns)
+        else:
+            self.articles["Preprint"] = None  # TODO add search for preprint of published articles
+            self.articles = self.articles[expected_columns]
 
 
 class PubMedClient:

tests/test_telegram.py

@@ -67,3 +67,11 @@ async def test_publish_many_papers(publisher, papers):
     message = await publisher.publish_papers(papers, preprints, today=None, spreadsheet_id=None)
 
     assert message.message_id is not None
+
+
+@pytest.mark.asyncio
+async def test_publish_empty_list(publisher, papers):
+    papers, preprints = publisher.format_papers([])
+    message = await publisher.publish_papers(papers, preprints, today=None, spreadsheet_id=None)
+
+    assert message.message_id is not None