add tests for flask api and the first /analyze endpoint

src/analyzer_engine/csv_analyzer_engine.py

@@ -28,4 +28,4 @@ def analyze_csv(
             csv_list = list(csv.reader(csv_file))
             csv_dict = {header: list(map(str, values)) for header, *values in zip(*csv_list)}
             analyzer_results = self.analyze_dict(csv_dict, language, keys_to_skip)
-            return analyzer_results
+            return list(analyzer_results)

app.py → src/app.py

@@ -1,6 +1,7 @@
 import json
 import logging
 import os
+import uuid
 from typing import Tuple
 
 from flask import Flask, request, jsonify, Response
@@ -36,26 +37,29 @@ def analyze() -> Tuple[str, int]:
                 if file.filename == '':
                     return jsonify({'error': 'No selected file'}), 400
 
-                filepath = f'uploads/{file.filename}'
+                filepath = f'uploads/{uuid.uuid4()}'
                 file.save(filepath)
                 self.logger.info(f"Successfully saved file: {filepath}")
 
-                analyzer_result_list = self.engine.analyze_csv(
+                analyzer_results = self.engine.analyze_csv(
                     csv_full_path=filepath,
                     language="en"
                 )
+                self.logger.debug(f"Analyzed file with results: {analyzer_results}")
+                os.remove(filepath)
+                self.logger.info(f"Successfully removed file: {filepath}")
 
-                resp = {}
-                for result in analyzer_result_list:
-                    resp['key'] = result.key
-                    resp['value'] = result.value
-                    resp['recognizer_results'] = json.dumps(
-                        result.recognizer_results,
-                        default=lambda o: o.to_dict(),
-                        sort_keys=True,
-                    )
+                analyzer_results_dict = {}
+                for a in analyzer_results:
+                    recognizer_results = []
+                    for r in a.recognizer_results:
+                        recognizer_results.append([o.to_dict() for o in r])
+                    analyzer_results_dict[a.key] = {
+                        "value": a.value,
+                        "recognizer_results": recognizer_results
+                    }
 
-                return jsonify(resp), 200
+                return jsonify(analyzer_results_dict), 200
             except Exception as e:
                 self.logger.error(
                     f"A fatal error occurred during execution of "

tests/app_test.py

@@ -0,0 +1,42 @@
+import pytest
+import json
+
+from app import Server
+
+@pytest.fixture()
+def app():
+    app = Server().app
+    app.config.update({
+        "TESTING": True,
+    })
+
+    yield app
+
+
+@pytest.fixture()
+def client(app):
+    return app.test_client()
+
+
+def test_health(client):
+    response = client.get("/health")
+    assert response.status_code == 200
+
+
+def test_analyze_csv_file(client):
+    expected_response_id = {'value': ['1', '2', '3'], 'recognizer_results': [[], [], []]}
+
+    response = client.post("/analyze", data={
+        "file": open('./tests/analyzer_engine/sample_data.csv', 'rb'),
+    })
+
+    assert response.status_code == 200
+    data = json.loads(response.get_data(as_text=True))
+    # No PII in id
+    assert data['id'] == expected_response_id
+    # first row has no PII
+    assert data['comments']['recognizer_results'][0] == []
+    # second row has PII
+    assert data['comments']['recognizer_results'][1][0]['entity_type'] == 'US_DRIVER_LICENSE'
+    assert data['comments']['recognizer_results'][1][0]['start'] == 34
+    assert data['comments']['recognizer_results'][1][0]['end'] == 42