Make OpenAlex work

src/paperqa/clients/openalex.py

@@ -90,7 +90,6 @@ async def get_doc_details_from_openalex(
 
     if fields:
         params["select"] = ",".join(fields)
-
     response = await client.get(
         url, params=params, timeout=OPENALEX_API_REQUEST_TIMEOUT
     )
@@ -104,6 +103,7 @@ async def get_doc_details_from_openalex(
         raise DOINotFoundError("OpenAlex API returned a failed status for the query.")
 
     results_data = response_data
+
     if params.get("filter") is not None:
         results_data = results_data["results"]
         if len(results_data) == 0:
@@ -112,13 +112,18 @@ async def get_doc_details_from_openalex(
             )
         results_data = results_data[0]
 
+    # openalex keeps the DOI prefix on (we remove)
+    if results_data.get("doi"):
+        results_data["doi"] = results_data["doi"].removeprefix("https://doi.org/")
+
     if (
         doi is None
         and title
         and strings_similarity(results_data.get("title", ""), title)
         < title_similarity_threshold
     ):
         raise DOINotFoundError(f"OpenAlex results did not match for title {title!r}.")
+
     if doi and results_data.get("doi") != doi:
         raise DOINotFoundError(f"DOI {doi!r} not found in OpenAlex.")
 

tests/test_clients.py

@@ -21,7 +21,7 @@
 )
 from paperqa.clients.client_models import MetadataPostProcessor, MetadataProvider
 from paperqa.clients.journal_quality import JournalQualityPostProcessor
-from paperqa.clients.openalex import reformat_name
+from paperqa.clients.openalex import OpenAlexProvider, reformat_name
 from paperqa.clients.retractions import RetractionDataPostProcessor
 from paperqa.types import DocDetails
 
@@ -793,3 +793,24 @@ async def test_tricky_journal_quality_results(doi: str, score: int) -> None:
         assert (
             crossref_details.source_quality == score
         ), "Should have source quality data"
+
+
+@pytest.mark.vcr
+@pytest.mark.parametrize(
+    ("doi"),
+    [
+        ("10.1021/acs.jctc.5b00178"),
+    ],
+)
+@pytest.mark.asyncio
+async def test_does_openalex_work(doi: str):
+    async with httpx_aiohttp.HttpxAiohttpClient() as http_client:
+        openalex_client = DocMetadataClient(
+            http_client,
+            metadata_clients=[OpenAlexProvider],
+        )
+        openalex_details = await openalex_client.query(
+            doi=doi,
+            fields=["open_access"],
+        )
+        assert openalex_details, "Failed to query OpenAlex"