Downgrade to FastAPI 0.68.2

- Using this version of FastAPI gives us Starlette 0.14.2 which fixes a bug we faced
- Bug described at encode/starlette#1315
- Downgrading FastAPI means `UploadFile` is no longer supported. I've reverted to `File()` as an alternative solution, although `UploadFile` is probably easier to use, and I'd reimplement it in that way if we could upgrade in the future

operationsgateway_api/src/routes/ingest_data.py

@@ -1,6 +1,6 @@
 import logging
 
-from fastapi import APIRouter, UploadFile
+from fastapi import APIRouter, File
 
 from operationsgateway_api.src.data_encoding import DataEncoding
 from operationsgateway_api.src.hdf_handler import HDFDataHandler
@@ -21,12 +21,11 @@
 
 
 @router.post("/submit_hdf", response_description="")
-async def submit_hdf(file: UploadFile):
+async def submit_hdf(file: bytes = File(default=None)):
     log.info("Submitting CLF data in HDF file to be processed then stored in MongoDB")
-    log.debug("Filename: %s, Content: %s", file.filename, file.content_type)
+    #log.debug("Filename: %s, Content: %s", file.filename, file.content_type)
 
-    file_contents = await file.read()
-    hdf_file = HDFDataHandler.convert_to_hdf_from_request(file_contents)
+    hdf_file = HDFDataHandler.convert_to_hdf_from_request(file)
 
     record, waveforms, images = HDFDataHandler.extract_hdf_data(hdf_file=hdf_file)
     DataEncoding.encode_numpy_for_mongo(record)
@@ -81,12 +80,11 @@ async def submit_hdf(file: UploadFile):
     response_description="Upload a HDF file, extract the contents and store the data in"
     " MongoDB",
 )
-async def submit_hdf_basic(file: UploadFile):
+async def submit_hdf_basic(file: File(default=None)):
     log.info("Adding contents of attached file to MongoDB")
-    log.debug("Filename: %s, Content: %s", file.filename, file.content_type)
+    #log.debug("Filename: %s, Content: %s", file.filename, file.content_type)
 
-    file_contents = await file.read()
-    hdf_file = HDFDataHandler.convert_to_hdf_from_request(file_contents)
+    hdf_file = HDFDataHandler.convert_to_hdf_from_request(file)
 
     # TODO - do something with these images
     record, waveforms, images = HDFDataHandler.extract_hdf_data(hdf_file=hdf_file)