Fixed Docs.aadd giving bad docname after upgrade (#1025)

Author: jamesbraza

src/paperqa/docs.py

@@ -300,10 +300,13 @@ async def aadd(  # noqa: PLR0912
             ):
                 citation = f"Unknown, {os.path.basename(path)}, {datetime.now().year}"
 
-        docname = citation_to_docname(citation) if docname is None else docname
-        docname = self._get_unique_name(docname)
-
-        doc = Doc(docname=docname, citation=citation, dockey=dockey)
+        doc = Doc(
+            docname=self._get_unique_name(
+                citation_to_docname(citation) if docname is None else docname
+            ),
+            citation=citation,
+            dockey=dockey,
+        )
 
         # try to extract DOI / title from the citation
         if (doi is title is None) and parse_config.use_doc_details:
@@ -402,7 +405,7 @@ async def aadd(  # noqa: PLR0912
                 " to ignore this error."
             )
         if await self.aadd_texts(texts, doc, all_settings, embedding_model):
-            return docname
+            return doc.docname
         return None
 
     def add_texts(

tests/test_paperqa.py

@@ -934,17 +934,17 @@ async def test_repeat_keys(stub_data_dir) -> None:
 @pytest.mark.asyncio
 async def test_pdf_reader_w_no_match_doc_details(stub_data_dir: Path) -> None:
     docs = Docs()
-    await docs.aadd(
+    docname = await docs.aadd(
         stub_data_dir / "paper.pdf",
         "Wellawatte et al, XAI Review, 2023",
     )
+    (doc_details,) = docs.docs.values()
+    assert doc_details.docname == docname, "Added name should match between details"
     # doc will be a DocDetails object, but nothing can be found
     # thus, we retain the prior citation data
     assert (
-        next(iter(docs.docs.values())).citation == "Wellawatte et al, XAI Review, 2023"
-    )
-    assert (
-        next(iter(docs.docs.values())).formatted_citation
+        doc_details.citation
+        == doc_details.formatted_citation
         == "Wellawatte et al, XAI Review, 2023"
     ), "Formatted citation should be the same when no metadata is found."
 
@@ -1016,7 +1016,7 @@ async def test_partly_embedded_texts(defer_embeddings: bool) -> None:
 @pytest.mark.asyncio
 async def test_pdf_reader_match_doc_details(stub_data_dir: Path) -> None:
     docs = Docs()
-    await docs.aadd(
+    docname = await docs.aadd(
         stub_data_dir / "paper.pdf",
         "Wellawatte et al, A Perspective on Explanations of Molecular Prediction"
         " Models, XAI Review, 2023",
@@ -1027,7 +1027,8 @@ async def test_pdf_reader_match_doc_details(stub_data_dir: Path) -> None:
         },  # Limit to only crossref since s2 is too flaky
         fields=["author", "journal", "citation_count"],
     )
-    doc_details = next(iter(docs.docs.values()))
+    (doc_details,) = docs.docs.values()
+    assert doc_details.docname == docname, "Added name should match between details"
     # Crossref is non-deterministic in its ordering for results
     # thus we need to capture both possible dockeys
     assert doc_details.dockey in {"d7763485f06aabde", "5300ef1d5fb960d7"}