🚧 respect threshold in cli and api

README.md

@@ -25,6 +25,11 @@ The implementation uses a pre-populated redis cache which can fallback to other
     juditha lookup "jane doe"
     "Jane Doe"
 
+To match more fuzzy, reduce the threshold (default 0.97):
+
+    juditha lookup "doe, jane" --threshold 0.5
+    "Jane Doe"
+
 ## data import
 
 ### from ftm entities
@@ -49,7 +54,8 @@ Following the [`nomenklatura`](https://github.com/opensanctions/nomenklatura) sp
 from juditha import lookup
 
 assert lookup("jane doe") == "Jane Doe"
-assert lookup("foo") is None
+assert lookup("doe, jane") is None
+assert lookup("doe, jane", threshold=0.5) == "Jane Doe"
 ```
 
 ## run as api
@@ -60,16 +66,16 @@ assert lookup("foo") is None
 
 Just do head requests to check if a name is known:
 
-    curl -I "http://localhost:8000/Alice"
+    curl -I "http://localhost:8000/jane%20doe"
     HTTP/1.1 200 OK
 
     curl -I "http://localhost:8000/John"
     HTTP/1.1 404 Not Found
 
-Do an actual request to get the canonized name:
+An actual request returns the canonized name:
 
-    curl "http://localhost:8000/alice"
-    Alice
+    curl "http://localhost:8000/doe,%20jane?threshold=0.5"
+    Jane Doe
 
 
 ## settings

juditha/api.py

@@ -14,23 +14,25 @@
 
 
 @app.get("/_classify/{q}")
-async def api_classify(q: str) -> str:
+async def api_classify(q: str) -> Response:
     schema = classify(q)
     if schema is None:
         return Response("404", status_code=404)
     return Response(schema)
 
 
 @app.get("/{q}")
-async def api_lookup(q: str, threshold: float | None = settings.FUZZY_THRESHOLD) -> str:
+async def api_lookup(
+    q: str, threshold: float | None = settings.FUZZY_THRESHOLD
+) -> Response:
     name = lookup(q, threshold=threshold)
     if name is None:
         return Response("404", status_code=404)
     return Response(name)
 
 
 @app.head("/{q}")
-async def api_head(q: str, threshold: float | None = settings.FUZZY_THRESHOLD) -> None:
+async def api_head(q: str, threshold: float | None = settings.FUZZY_THRESHOLD) -> int:
     name = lookup(q, threshold=threshold)
     if name is None:
         raise HTTPException(404)

juditha/cli.py

@@ -45,7 +45,7 @@ def load_dataset(
     with_schema: Annotated[
         bool, typer.Option(..., help="Include schemata for classifier")
     ] = False,
-) -> int:
+):
     try:
         res = io.load_dataset(uri, with_schema=with_schema)
         success(f"Imported {res} names.")
@@ -59,7 +59,7 @@ def load_catalog(
     with_schema: Annotated[
         bool, typer.Option(..., help="Include schemata for classifier")
     ] = False,
-) -> int:
+):
     try:
         res = io.load_catalog(uri, with_schema=with_schema)
         success(f"Imported {res} names.")
@@ -68,9 +68,14 @@ def load_catalog(
 
 
 @cli.command("lookup")
-def cli_lookup(value: str):
+def cli_lookup(
+    value: str,
+    threshold: Annotated[
+        float, typer.Option(..., help="Fuzzy threshold")
+    ] = settings.FUZZY_THRESHOLD,
+):
     try:
-        result = lookup(value)
+        result = lookup(value, threshold=threshold)
         if result is not None:
             print(result)
         else:

tests/fixtures/names.txt

@@ -0,0 +1,2 @@
+Jane Doe
+Alice

tests/test_cli.py

@@ -0,0 +1,27 @@
+from pathlib import Path
+
+from typer.testing import CliRunner
+
+from juditha.cli import cli
+
+runner = CliRunner()
+
+
+def test_cli(fixtures_path: Path):
+    runner.invoke(cli, ["load", "-i", fixtures_path / "names.txt"])
+    res = runner.invoke(cli, ["lookup", "Jane Doe"])
+    assert res.exit_code == 0
+    assert res.output.strip() == "Jane Doe"
+
+    res = runner.invoke(cli, ["lookup", "doe, jane"])
+    assert res.exit_code == 0
+    assert "not found" in res.output
+    res = runner.invoke(cli, ["lookup", "doe, jane", "--threshold", "0.5"])
+    assert res.exit_code == 0
+    assert res.output.strip() == "Jane Doe"
+
+    res = runner.invoke(
+        cli,
+        ["load", "-i", fixtures_path / "eu_authorities.ftm.json", "--from-entities"],
+    )
+    assert res.exit_code == 0