[WIP] add api endpoints to support anonymous data check

.devcontainer/devcontainer.json

@@ -35,6 +35,7 @@
         "editor.tabSize": 4,
         "editor.formatOnSave": true,
         "python.formatting.provider": "black",
+        "python.formatting.black": true,
         "python.envFile": "${workspaceFolder}/.env",
         "editor.codeActionsOnSave": {
           "source.organizeImports": true

.devcontainer/requirements.txt

@@ -5,4 +5,8 @@ pandas==1.5.3
 pandera==0.14.5
 ipykernel
 openpyxl
-pytest
+pytest
+python-multipart
+uvicorn
+fastapi
+httpx

config.py

@@ -1,16 +1,16 @@
 # path to original/raw NAICS excel file
-NAICS_EXCEL_PATH = "./data/naics/raw/2-6 digit_2022_Codes.xlsx"
+NAICS_EXCEL_PATH = "data/naics/raw/2-6 digit_2022_Codes.xlsx"
 # path to parsed/filtered naics codes file
-NAICS_CSV_PATH = "./data/naics/processed/2022_codes.csv"
+NAICS_CSV_PATH = "data/naics/processed/2022_codes.csv"
 # column header text containing naics code
 NAICS_CODE_COL = "2022 NAICS US   Code"
 # column header text containing naics title/description
 NAICS_TITLE_COL = "2022 NAICS US Title"
 
 # path to original/raw NAICS zip file
-CENSUS_RAW_ZIP_PATH = "./data/census/raw/CensusFlatFile2022.zip"
+CENSUS_RAW_ZIP_PATH = "data/census/raw/CensusFlatFile2022.zip"
 # path to parsed/filtered naics codes file
-CENSUS_PROCESSED_CSV_PATH = "./data/census/processed/Census2022.processed.csv"
+CENSUS_PROCESSED_CSV_PATH = "data/census/processed/Census2022.processed.csv"
 # census file col indexes
 CENSUS_STATE_COL_INDEX = 2
 CENSUS_COUNTY_COL_INDEX = 3

src/__init__.py

@@ -0,0 +1,5 @@
+import os
+import sys
+
+CURR_DIR = os.path.dirname(os.path.abspath(__file__)) # noqa: E402
+sys.path.append(CURR_DIR) # noqa: E402

src/api/__init__.py

@@ -0,0 +1,9 @@
+import os
+import sys
+
+CURR_DIR = os.path.dirname(os.path.abspath(__file__)) # noqa: E402
+PARENT_DIR = os.path.dirname(CURR_DIR) # noqa: E402
+ROOT_DIR = os.path.dirname(PARENT_DIR) # noqa: E402
+sys.path.append(CURR_DIR) # noqa: E402
+sys.path.append(PARENT_DIR) # noqa: E402
+sys.path.append(ROOT_DIR) # noqa: E402

src/api/main.py

@@ -0,0 +1,44 @@
+from fastapi import FastAPI, Request
+from fastapi.responses import JSONResponse
+from routers import validate_router
+
+from util.global_data import read_geoids, read_naics_codes
+from validator.create_schemas import get_schemas
+
+# HOWTO
+# run: uvicorn main:app --host 0.0.0.0 --port 8080 --reload
+
+app = FastAPI()
+# get our schems
+app.schemas = get_schemas(read_naics_codes(), read_geoids())
+
+
+# custom exception handlers
+@app.exception_handler(Exception)
+async def api_exception_handler(request: Request, err: Exception):
+    """
+    handle exception/error that is triggered by incorrect data format
+
+    Args:
+        request (Request): HTTP request
+        err (Exception): the actual exception object
+
+    Returns:
+        returns JSON Response with status code and message
+    """
+    error_code = 400
+    error_msg = str(err)
+    if type(err) is ValueError:
+        error_code = 422
+    elif type(err) is AttributeError:
+        error_code = 500
+        error_msg = f"Internal Server Error. Type: {str(type(err))}"
+
+    return JSONResponse(
+        status_code=error_code,
+        content=[{"response": error_msg}],
+    )
+
+
+# add anonymous validation endpoints
+app.include_router(validate_router, prefix="/v1")

src/api/routers/__init__.py

@@ -0,0 +1,3 @@
+__all__ = ["validate_router"]
+
+from .validate import router as validate_router

src/api/routers/validate.py

@@ -0,0 +1,44 @@
+from fastapi import APIRouter, Request, UploadFile
+from fastapi.responses import JSONResponse
+
+from validator.create_schemas import validate_data_list, validate_raw_csv
+
+# validate router
+router = APIRouter()
+
+
+@router.post("/validate/")
+async def validate(request: Request, payload: dict):
+    """
+    handle validate endpoint
+
+    Args:
+        request (Request): request object
+        payload (dict): data received from user
+
+    Returns:
+        JSON response
+    """
+    phase1_schema = request.app.schemas[0]
+    phase2_schema = request.app.schemas[1]
+    response = validate_data_list(phase1_schema, phase2_schema, payload)
+    return JSONResponse(status_code=200, content=response)
+
+
+@router.post("/upload/")
+async def upload(request: Request, file: UploadFile):
+    """
+    handle upload request
+
+    Args:
+        request (Request): request object
+        file (UploadFile): uploaded file
+
+    Returns:
+        JSON response
+    """
+    contents = file.file.read()
+    phase1_schema = request.app.schemas[0]
+    phase2_schema = request.app.schemas[1]
+    response = validate_raw_csv(phase1_schema, phase2_schema, contents)
+    return JSONResponse(status_code=200, content=response)

src/tests/__init__.py

@@ -1,6 +1,10 @@
 import os
 import sys
 
-ROOT_DIR = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
-
-sys.path.append(os.path.join(ROOT_DIR, "validator"))
+CURR_DIR = os.path.dirname(os.path.abspath(__file__))  # noqa: E402
+PARENT_DIR = os.path.dirname(CURR_DIR)  # noqa: E402
+ROOT_DIR = os.path.dirname(PARENT_DIR)  # noqa: E402
+sys.path.append(CURR_DIR)  # noqa: E402
+sys.path.append(PARENT_DIR)  # noqa: E402
+sys.path.append(ROOT_DIR)  # noqa: E402
+sys.path.append(os.path.join(ROOT_DIR, "validator"))