Create DataFrame object when field value is None (#79)

* fix bug

* Update CHANGELOG.md

Co-authored-by: Troy Raen <[email protected]>

* implement requested changes

* update CHANGELOD

---------

Co-authored-by: Troy Raen <[email protected]>

Author: hernandezc1

Diff for: CHANGELOG.md

@@ -10,6 +10,12 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/)
 
 ## \[Unreleased\]
 
+### Fixed
+
+- Fix issue #76. Make `Alert.dataframe` succeed even when there are no previous sources.
+
+## \[v0.3.12\] - 2025-03-07
+
 ### Added
 
 - Data listings and references for LSST

Diff for: pittgoogle/alert.py

@@ -298,9 +298,9 @@ def dataframe(self) -> "pd.DataFrame":
         import pandas as pd  # always lazy-load pandas. it hogs memory on cloud functions and run
 
         # sources and previous sources are expected to have the same fields
-        sources_df = pd.DataFrame([self.get("source")] + self.get("prv_sources", []))
+        sources_df = pd.DataFrame([self.get("source")] + (self.get("prv_sources") or []))
         # sources and forced sources may have different fields
-        forced_df = pd.DataFrame(self.get("prv_forced_sources", []))
+        forced_df = pd.DataFrame(self.get("prv_forced_sources") or [])
 
         # use nullable integer data type to avoid converting ints to floats
         # for columns in one dataframe but not the other