Distopia22's picture
Add automatic PII removal during file extraction
fd20bd2
from fastapi.testclient import TestClient
from src.main import app
from io import BytesIO
client = TestClient(app)
def test_coding_endpoint():
"""Test existing text input endpoint"""
provider_notes = {
"provider_notes": "Patient has a fever and cough."
}
response = client.post("/api/coding", json=provider_notes)
assert response.status_code == 200
data = response.json()
assert "cpt_codes" in data
assert "icd_codes" in data
assert isinstance(data["cpt_codes"], list)
assert isinstance(data["icd_codes"], list)
def test_file_upload_endpoint():
"""Test new file upload endpoint"""
file_content = b"Patient John Doe presents with acute bronchitis. Cough for 5 days, productive with yellow sputum. Lung exam reveals diffuse wheezing."
files = {
"file": ("provider_notes.txt", BytesIO(file_content), "text/plain")
}
response = client.post("/api/upload-file", files=files)
assert response.status_code == 200
data = response.json()
assert data["success"] is True
assert data["filename"] == "provider_notes.txt"
assert data["extracted_text_length"] > 0
assert "pii_removed" in data
assert "pii_count" in data
assert "cpt_codes" in data
assert "icd_codes" in data
def test_file_upload_with_pii():
"""Test file upload with PII - should be removed"""
file_content = b"""
Patient: John Doe
DOB: 01/15/1980
Phone: 555-123-4567
Address: 123 Main St, New York, NY
Chief Complaint: Chest pain
History: Patient presents with acute chest pain...
"""
files = {
"file": ("notes_with_pii.txt", BytesIO(file_content), "text/plain")
}
response = client.post("/api/upload-file", files=files)
assert response.status_code == 200
data = response.json()
# PII should be detected and removed
assert data["pii_removed"] is True
assert data["pii_count"] > 0
def test_file_upload_invalid_extension():
"""Test file upload with invalid file type"""
file_content = b"Some content"
files = {
"file": ("document.pdf", BytesIO(file_content), "application/pdf")
}
response = client.post("/api/upload-file", files=files)
assert response.status_code == 400
def test_file_upload_empty_file():
"""Test file upload with empty file"""
file_content = b""
files = {
"file": ("empty.txt", BytesIO(file_content), "text/plain")
}
response = client.post("/api/upload-file", files=files)
assert response.status_code == 400