Introduce option to not validate dois

.github/workflows/python-check-pip.yml

@@ -19,7 +19,7 @@ jobs:
         python-version: ["3.10", "3.11", "3.12"]
 
     steps:
-    # - uses: actions/checkout@v4
+    - uses: actions/checkout@v4
     - name: Set up Python ${{ matrix.python-version }}
       uses: actions/setup-python@v3
       with:
@@ -28,6 +28,7 @@ jobs:
       run: |
         python -m pip install ash-williams
         python -m pip freeze
-    - name: Literally just invoke it # Consider a more elaborate test
+        ls
+    - name: Barebones demo script
       run: |
-        python -c "from ash import Paper, RetractionDatabase;print(ash.Paper)"
+        python ./test/demo.py

ash/ash.py

@@ -209,16 +209,32 @@ def from_path(cls, path: Path | str, mime_type: str | None = None) -> "Paper":
         with path.open("rb") as stream:
             return cls(stream, mime_type)
 
-    def report(self, db: RetractionDatabase | Path | str) -> dict[str, Any]:
+    def report(
+        self,
+        db: RetractionDatabase | Path | str,
+        validate_dois: bool = True,
+    ) -> dict[str, Any]:
         if isinstance(db, (Path, str)):
             db = RetractionDatabase(db)
-        all_dois = {
+        dois_report = self._generate_dois_report(db, validate=validate_dois)
+        zombie_report = self._generate_zombie_report(db)
+        return {"dois": dois_report, "zombies": zombie_report}
+
+    def _generate_dois_report(
+        self, db: RetractionDatabase, validate: bool
+    ) -> dict[str, Any]:
+        if not validate:
+            return {doi: {"Retracted": (doi in db.dois)} for doi in self.dois}
+
+        return {
             doi: {
                 "DOI is valid": DOI(doi).exists(),
                 "Retracted": (doi in db.dois),
             }
             for doi in self.dois
         }
+
+    def _generate_zombie_report(self, db: RetractionDatabase) -> list[dict[str, Any]]:
         zombies = sorted([doi for doi in self.dois if doi in db.dois])
         zombie_report = [
             {
@@ -230,7 +246,7 @@ def report(self, db: RetractionDatabase | Path | str) -> dict[str, Any]:
             for doi in zombies
             for record in db.data[doi]
         ]
-        return {"dois": all_dois, "zombies": zombie_report}
+        return zombie_report
 
     @classmethod
     def register_handler(

test/demo.py

@@ -0,0 +1,10 @@
+import ash
+
+TEXT = """
+You can find this in the sporting goods department. That's right, this sweet baby was
+made in Grand Rapids, Michigan. Retails for about doi: 10.10995/walnutstock.
+That's right. "Short Smart: Shop S-Mart!"
+"""
+db = ash.RetractionDatabase("./test/mock/rw_database.csv")
+paper = ash.Paper(TEXT, mime_type="text/plain")
+print(paper.report(db, validate_dois=False))

test/test_ash.py

@@ -149,6 +149,15 @@ def test_unclear_response_from_server(self, fake_db, mock_http, tmpdir):
             "Retracted": True,
         }
 
+    def test_no_socket_hits_with_no_validation(self, fake_db):
+        text = """
+        You can find this in the sporting goods department. That's right, this sweet
+        baby was made in Grand Rapids, Michigan. Retails for about
+        doi:10.10995/walnutstock. That's right. "Short Smart: Shop S-Mart!"
+        """
+        paper = Paper(text, mime_type="text/plain")
+        print(paper.report(fake_db, validate_dois=False))
+
 
 class TestMIMEBehavior:
     @pytest.mark.parametrize(