bugfix/fix file data serialization and add unit tests (#34)

Author: rbiseck3

CHANGELOG.md

@@ -1,3 +1,7 @@
+## 0.0.15
+
+* **Bugfix for file data serialization**
+
 ## 0.0.14
 
 * **Add support for batch file data**

requirements/cli.txt

@@ -32,7 +32,7 @@ idna==3.10
     # via anyio
 importlib-metadata==8.5.0
     # via opentelemetry-api
-marshmallow==3.23.1
+marshmallow==3.23.2
     # via dataclasses-json
 mypy-extensions==1.0.0
     # via typing-inspect
@@ -84,15 +84,15 @@ packaging==24.2
     #   opentelemetry-instrumentation
 pandas==2.2.3
     # via unstructured-ingest
-protobuf==5.29.1
+protobuf==5.29.2
     # via
     #   googleapis-common-protos
     #   opentelemetry-proto
-pydantic==2.10.3
+pydantic==2.10.4
     # via
     #   fastapi
     #   unstructured-ingest
-pydantic-core==2.27.1
+pydantic-core==2.27.2
     # via pydantic
 python-dateutil==2.9.0.post0
     # via
@@ -122,7 +122,7 @@ typing-inspect==0.9.0
     # via dataclasses-json
 tzdata==2024.2
     # via pandas
-unstructured-ingest==0.3.10
+unstructured-ingest==0.3.11
     # via -r ./cli.in
 uvicorn==0.34.0
     # via -r ./cli.in

requirements/constraints.txt

@@ -1 +1 @@
-unstructured-ingest>=0.3.10
+unstructured-ingest>=0.3.1

requirements/lint.txt

@@ -34,7 +34,7 @@ pyflakes==3.2.0
     # via
     #   autoflake
     #   flake8
-ruff==0.8.3
+ruff==0.8.4
     # via -r ./lint.in
 tomli==2.2.1
     # via

requirements/test.in

@@ -1 +1,2 @@
 pytest
+httpx

requirements/test.txt

@@ -1,7 +1,25 @@
 # This file was autogenerated by uv via the following command:
 #    uv pip compile ./test.in --output-file ./test.txt --no-strip-extras --python-version 3.10
+anyio==4.7.0
+    # via httpx
+certifi==2024.12.14
+    # via
+    #   httpcore
+    #   httpx
 exceptiongroup==1.2.2
-    # via pytest
+    # via
+    #   anyio
+    #   pytest
+h11==0.14.0
+    # via httpcore
+httpcore==1.0.7
+    # via httpx
+httpx==0.28.1
+    # via -r ./test.in
+idna==3.10
+    # via
+    #   anyio
+    #   httpx
 iniconfig==2.0.0
     # via pytest
 packaging==24.2
@@ -10,5 +28,9 @@ pluggy==1.5.0
     # via pytest
 pytest==8.3.4
     # via -r ./test.in
+sniffio==1.3.1
+    # via anyio
 tomli==2.2.1
     # via pytest
+typing-extensions==4.12.2
+    # via anyio

test/api/__init__.py

@@ -0,0 +1,125 @@
+from pathlib import Path
+from typing import Any, Optional
+
+import pytest
+from fastapi.testclient import TestClient
+from pydantic import BaseModel
+from unstructured_ingest.v2.interfaces import BatchFileData, BatchItem, FileData, SourceIdentifiers
+
+from unstructured_platform_plugins.etl_uvicorn.api_generator import (
+    EtlApiException,
+    UsageData,
+    wrap_in_fastapi,
+)
+from unstructured_platform_plugins.schema.filedata_meta import FileDataMeta
+
+
+class InvokeResponse(BaseModel):
+    usage: list[UsageData]
+    status_code: int
+    filedata_meta: FileDataMeta
+    status_code_text: Optional[str] = None
+    output: Optional[Any] = None
+
+    def generic_validation(self):
+        assert self.status_code == 200
+        assert not self.status_code_text
+
+
+mock_file_data = [
+    FileData(
+        identifier="mock file data",
+        connector_type="CON",
+        source_identifiers=SourceIdentifiers(filename="n", fullpath="n"),
+    ),
+    BatchFileData(
+        identifier="mock file data", connector_type="CON", batch_items=[BatchItem(identifier="bid")]
+    ),
+]
+
+
+@pytest.mark.parametrize(
+    "file_data", mock_file_data, ids=[type(fd).__name__ for fd in mock_file_data]
+)
+def test_async_sample_function(file_data):
+    from test.assets.async_typed_dict_response import async_sample_function as test_fn
+
+    client = TestClient(wrap_in_fastapi(func=test_fn, plugin_id="mock_plugin"))
+
+    post_body = {"file_data": file_data.model_dump(), "content": {"a": 1, "b": 2}}
+    resp = client.post("/invoke", json=post_body)
+    resp_content = resp.json()
+    invoke_response = InvokeResponse.model_validate(resp_content)
+    invoke_response.generic_validation()
+    output = invoke_response.output
+    assert isinstance(output, dict)
+    assert output == {"response": {"a_out": 1, "b_out": 2}}
+
+
+@pytest.mark.parametrize(
+    "file_data", mock_file_data, ids=[type(fd).__name__ for fd in mock_file_data]
+)
+def test_dataclass_response(file_data):
+    from test.assets.dataclass_response import sample_function_with_path as test_fn
+
+    client = TestClient(wrap_in_fastapi(func=test_fn, plugin_id="mock_plugin"))
+    current_path = Path(__file__)
+
+    post_body = {"file_data": file_data.model_dump(), "c": 1, "b": "2", "a": str(current_path)}
+    resp = client.post("/invoke", json=post_body)
+    resp_content = resp.json()
+    invoke_response = InvokeResponse.model_validate(resp_content)
+    invoke_response.generic_validation()
+    output = invoke_response.output
+    assert isinstance(output, dict)
+    assert output == {
+        "t": "PosixPath",
+        "exists": True,
+        "resolved": str(current_path.resolve()),
+        "b": "2",
+        "c": 1,
+    }
+
+
+@pytest.mark.parametrize(
+    "file_data", mock_file_data, ids=[type(fd).__name__ for fd in mock_file_data]
+)
+def test_empty_input_and_output(file_data):
+    from test.assets.empty_input_and_response import SampleClass as TestClass
+
+    test_class = TestClass()
+    client = TestClient(wrap_in_fastapi(func=test_class.do_nothing, plugin_id="mock_plugin"))
+
+    resp = client.post("/invoke", json={"file_data": file_data.model_dump()})
+    resp_content = resp.json()
+    invoke_response = InvokeResponse.model_validate(resp_content)
+    invoke_response.generic_validation()
+    output = invoke_response.output
+    assert not output
+
+
+@pytest.mark.parametrize(
+    "file_data", mock_file_data, ids=[type(fd).__name__ for fd in mock_file_data]
+)
+def test_filedata_meta(file_data):
+    from test.assets.filedata_meta import Input
+    from test.assets.filedata_meta import process_input as test_fn
+
+    client = TestClient(wrap_in_fastapi(func=test_fn, plugin_id="mock_plugin"))
+
+    post_body = {"file_data": file_data.model_dump(), "i": Input(m=15).model_dump()}
+    resp = client.post("/invoke", json=post_body)
+    resp_content = resp.json()
+    invoke_response = InvokeResponse.model_validate(resp_content)
+    invoke_response.generic_validation()
+    filedata_meta = invoke_response.filedata_meta
+    assert len(filedata_meta.new_records) == 15
+    assert filedata_meta.terminate_current
+    assert not invoke_response.output
+
+
+def test_improper_function():
+    from test.assets.improper_function import sample_improper_function as test_fn
+
+    with pytest.raises(EtlApiException):
+        TestClient(wrap_in_fastapi(func=test_fn, plugin_id="mock_plugin"))

test/api/test_api.py

@@ -1,4 +1,6 @@
-from typing import Any, TypedDict
+from typing import Any
+
+from typing_extensions import TypedDict
 
 
 class SampleFunctionResponse(TypedDict):

test/assets/async_typed_dict_response.py

@@ -1,8 +1,8 @@
 import math
-from typing import Optional
+from typing import Optional, Union
 
 from pydantic import BaseModel
-from unstructured_ingest.v2.interfaces import FileData
+from unstructured_ingest.v2.interfaces import BatchFileData, FileData, SourceIdentifiers
 
 from unstructured_platform_plugins.schema import FileDataMeta, NewRecord
 
@@ -15,17 +15,21 @@ class Output(BaseModel):
     n: float
 
 
-def process_input(i: Input, file_data: FileData, filedata_meta: FileDataMeta) -> Optional[Output]:
+def process_input(
+    i: Input, file_data: Union[FileData, BatchFileData], filedata_meta: FileDataMeta
+) -> Optional[Output]:
     if i.m > 10:
         filedata_meta.terminate_current = True
         new_content = [
             NewRecord(
                 file_data=FileData(
-                    identifier=str(i.m + x), connector_type=file_data.connector_type
+                    identifier=str(i.m + x),
+                    connector_type=file_data.connector_type,
+                    source_identifiers=SourceIdentifiers(filename=f"{x}.txt", fullpath=f"{x}.txt"),
                 ),
                 contents=Output(n=float(i.m + x)),
             )
-            for x in range(5)
+            for x in range(i.m)
         ]
         filedata_meta.new_records.extend(new_content)
         return None