From 2f77206b8fa99c82171687d0866b41e1e0621284 Mon Sep 17 00:00:00 2001
From: Bertrand Benjamin <benjamin.bertrand@opytex.org>
Date: Sun, 18 Feb 2024 17:40:52 +0100
Subject: [PATCH] Feat: add Interseptor

---
 scripts/__init__.py               |  0
 scripts/intersept_not_valid.py    | 28 ++++++++++++++
 tests/__ini__.py                  |  0
 tests/test_intersept_not_valid.py | 63 +++++++++++++++++++++++++++++++
 4 files changed, 91 insertions(+)
 create mode 100644 scripts/__init__.py
 create mode 100644 scripts/intersept_not_valid.py
 create mode 100644 tests/__ini__.py
 create mode 100644 tests/test_intersept_not_valid.py

diff --git a/scripts/__init__.py b/scripts/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/scripts/intersept_not_valid.py b/scripts/intersept_not_valid.py
new file mode 100644
index 0000000..db0b41f
--- /dev/null
+++ b/scripts/intersept_not_valid.py
@@ -0,0 +1,28 @@
+from collections.abc import Callable
+
+import pandas as pd
+from pydantic import BaseModel, ValidationError
+
+
+class Interseptor:
+    def __init__(self, model: BaseModel):
+        self.model = model
+        self.not_valid_rows = []
+
+    def __call__(self, func: Callable[..., pd.DataFrame]):
+        def wrapped(*args, **kwrds):
+            res = func(*args, **kwrds)
+            df_dict = res.to_dict(orient="records")
+            valid_rows = []
+            for i, r in enumerate(df_dict):
+                try:
+                    self.model(**r)
+                except ValidationError:
+                    r["InterseptorOrigin"] = func.__name__
+                    r["InterseptorIndex"] = i
+                    self.not_valid_rows.append(r)
+                else:
+                    valid_rows.append(r)
+            return pd.DataFrame.from_records(valid_rows)
+
+        return wrapped
diff --git a/tests/__ini__.py b/tests/__ini__.py
new file mode 100644
index 0000000..e69de29
diff --git a/tests/test_intersept_not_valid.py b/tests/test_intersept_not_valid.py
new file mode 100644
index 0000000..d91098c
--- /dev/null
+++ b/tests/test_intersept_not_valid.py
@@ -0,0 +1,63 @@
+import random
+
+import pandas as pd
+import pytest
+from pydantic import BaseModel
+
+from scripts.intersept_not_valid import Interseptor
+
+
+class FakeModel(BaseModel):
+    name: str
+    age: int
+
+
+def test_init_composed():
+    interceptor = Interseptor(FakeModel)
+
+    def df_generator(nrows=3):
+        records = [{"name": "plop", "age": random.randint(1, 50)} for _ in range(nrows)]
+        return pd.DataFrame.from_records(records)
+
+    df_generator_val = interceptor(df_generator)
+
+    df = df_generator_val(3)
+    assert len(df) == 3
+    assert interceptor.not_valid_rows == []
+
+
+def test_init_decorator():
+    interceptor = Interseptor(FakeModel)
+
+    @interceptor
+    def df_generator(nrows=3):
+        records = [{"name": "plop", "age": random.randint(1, 50)} for _ in range(nrows)]
+        return pd.DataFrame.from_records(records)
+
+    df = df_generator(3)
+    assert len(df) == 3
+    assert interceptor.not_valid_rows == []
+
+
+def test_intersept_not_valid():
+    interceptor = Interseptor(FakeModel)
+
+    @interceptor
+    def df_generator():
+        records = [
+            {"name": "plop", "age": 12},
+            {"name": "hop", "age": "ui"},
+            {"name": "pipo", "age": 12},
+        ]
+        return pd.DataFrame.from_records(records)
+
+    df = df_generator()
+    assert len(df) == 2
+    assert interceptor.not_valid_rows == [
+        {
+            "name": "hop",
+            "age": "ui",
+            "InterseptorOrigin": "df_generator",
+            "InterseptorIndex": 1,
+        }
+    ]