Merge pull request #77 from ai-cfia/76-enhance-datastore-include-mode…

…l-in-json

Fixes #76: Enhance Data Storage: Including Model Details in the result.json file

README.md

@@ -83,6 +83,11 @@ backend to function, you will need to add the missing values:
   accepted by the backend
 - **NACHET_VALID_DIMENSION**: Contains the valid dimensions for an image to be
   accepted in the backend.
+- **NACHET_MAX_CONTENT_LENGTH**: Set the maximum size of the file that can be
+  uploaded to the backend. Needs to be the same size as the
+  `client_max_body_size`
+  [value](https://github.com/ai-cfia/howard/blob/dedee069f051ba743122084fcb5d5c97c2499359/kubernetes/aks/apps/nachet/base/nachet-ingress.yaml#L13)
+  set from the deployment in Howard.
 
 #### DEPRECATED
 

app.py

@@ -72,19 +72,6 @@ class MaxContentLengthWarning(APIWarnings):
 
 NACHET_DATA = os.getenv("NACHET_DATA")
 
-Model = namedtuple(
-    'Model',
-    [
-        'entry_function',
-        'name',
-        'endpoint',
-        'api_key',
-        'inference_function',
-        'content_type',
-        'deployment_platform',
-    ]
-)
-
 try:
     VALID_EXTENSION = json.loads(os.getenv("NACHET_VALID_EXTENSION"))
     VALID_DIMENSION = json.loads(os.getenv("NACHET_VALID_DIMENSION"))
@@ -115,6 +102,7 @@ class MaxContentLengthWarning(APIWarnings):
     [
         'request_function',
         'name',
+        'version',
         'endpoint',
         'api_key',
         'content_type',
@@ -371,7 +359,9 @@ async def inference_request():
             container_client, folder_name, image_bytes, hash_value
         )
 
-        for idx, model in enumerate(pipelines_endpoints.get(pipeline_name)):
+        pipeline = pipelines_endpoints.get(pipeline_name)
+
+        for idx, model in enumerate(pipeline):
             print(f"Entering {model.name.upper()} model") # TODO: Transform into logging
             result_json = await model.request_function(model, cache_json_result[idx])
             cache_json_result.append(result_json)
@@ -383,7 +373,7 @@ async def inference_request():
             cache_json_result[-1], imageDims, area_ratio, color_format
         )
 
-        result_json_string = json.dumps(processed_result_json)
+        result_json_string = await record_model(pipeline, processed_result_json)
 
         # upload the inference results to the user's container as async task
         app.add_background_task(
@@ -455,6 +445,7 @@ async def test():
     m = Model(
         request_function["test"],
         "test_model1",
+        1,
         "http://localhost:8080/test_model1",
         "test_api_key",
         "application/json",
@@ -465,6 +456,12 @@ async def test():
 
     return CACHE["endpoints"], 200
 
+
+async def record_model(pipeline: namedtuple, result: list):
+    new_entry = [{"name": model.name, "version": model.version} for model in pipeline]
+    result[0]["models"] = new_entry
+    return json.dumps(result, indent=4)
+
 
 async def fetch_json(repo_URL, key, file_path):
     """
@@ -505,6 +502,7 @@ async def get_pipelines(connection_string, pipeline_blob_name, pipeline_version,
         m = Model(
             request_function.get(model.get("endpoint_name")),
             model.get("model_name"),
+            model.get("version"),
             # To protect sensible data (API key and model endpoint), we encrypt it when
             # it's pushed into the blob storage. Once we retrieve the data here in the
             # backend, we need to decrypt the byte format to recover the original

tests/test_inference_request.py

@@ -65,7 +65,8 @@ def test_inference_request_successful(self, mock_container):
             "score",
             "topN",
             "overlapping",
-            "overlappingIndices"
+            "overlappingIndices",
+            "models"
         }
 
         # Test the answers from inference_request