[AUR-408] Export logs to Excel (#23)

This CL implements:

- The logic to export log to Excel.
- Route the export logic in the UI.
- Demonstrate this functionality in `./examples/promptui` project.

tests/simple_pipeline.py

@@ -0,0 +1,43 @@
+import tempfile
+from typing import List
+
+from theflow import Node
+
+from kotaemon.base import BaseComponent
+from kotaemon.embeddings import AzureOpenAIEmbeddings
+from kotaemon.llms.completions.openai import AzureOpenAI
+from kotaemon.pipelines.retrieving import RetrieveDocumentFromVectorStorePipeline
+from kotaemon.vectorstores import ChromaVectorStore
+
+
+class Pipeline(BaseComponent):
+    vectorstore_path: str = str(tempfile.mkdtemp())
+    llm: Node[AzureOpenAI] = Node(
+        default=AzureOpenAI,
+        default_kwargs={
+            "openai_api_base": "https://test.openai.azure.com/",
+            "openai_api_key": "some-key",
+            "openai_api_version": "2023-03-15-preview",
+            "deployment_name": "gpt35turbo",
+            "temperature": 0,
+            "request_timeout": 60,
+        },
+    )
+
+    @Node.decorate(depends_on=["vectorstore_path"])
+    def retrieving_pipeline(self):
+        vector_store = ChromaVectorStore(self.vectorstore_path)
+        embedding = AzureOpenAIEmbeddings(
+            model="text-embedding-ada-002",
+            deployment="embedding-deployment",
+            openai_api_base="https://test.openai.azure.com/",
+            openai_api_key="some-key",
+        )
+
+        return RetrieveDocumentFromVectorStorePipeline(
+            vector_store=vector_store, embedding=embedding
+        )
+
+    def run_raw(self, text: str) -> str:
+        matched_texts: List[str] = self.retrieving_pipeline(text)
+        return self.llm("\n".join(matched_texts)).text[0]

tests/test_promptui.py

@@ -1,66 +1,14 @@
-import pytest
-
 from kotaemon.contribs.promptui.config import export_pipeline_to_config
+from kotaemon.contribs.promptui.export import export_from_dict
 from kotaemon.contribs.promptui.ui import build_from_dict
 
-
-@pytest.fixture()
-def simple_pipeline_cls(tmp_path):
-    """Create a pipeline class that can be used"""
-    from typing import List
-
-    from theflow import Node
-
-    from kotaemon.base import BaseComponent
-    from kotaemon.embeddings import AzureOpenAIEmbeddings
-    from kotaemon.llms.completions.openai import AzureOpenAI
-    from kotaemon.pipelines.retrieving import (
-        RetrieveDocumentFromVectorStorePipeline,
-    )
-    from kotaemon.vectorstores import ChromaVectorStore
-
-    class Pipeline(BaseComponent):
-        vectorstore_path: str = str(tmp_path)
-        llm: Node[AzureOpenAI] = Node(
-            default=AzureOpenAI,
-            default_kwargs={
-                "openai_api_base": "https://test.openai.azure.com/",
-                "openai_api_key": "some-key",
-                "openai_api_version": "2023-03-15-preview",
-                "deployment_name": "gpt35turbo",
-                "temperature": 0,
-                "request_timeout": 60,
-            },
-        )
-
-        @Node.decorate(depends_on=["vectorstore_path"])
-        def retrieving_pipeline(self):
-            vector_store = ChromaVectorStore(self.vectorstore_path)
-            embedding = AzureOpenAIEmbeddings(
-                model="text-embedding-ada-002",
-                deployment="embedding-deployment",
-                openai_api_base="https://test.openai.azure.com/",
-                openai_api_key="some-key",
-            )
-
-            return RetrieveDocumentFromVectorStorePipeline(
-                vector_store=vector_store, embedding=embedding
-            )
-
-        def run_raw(self, text: str) -> str:
-            matched_texts: List[str] = self.retrieving_pipeline(text)
-            return self.llm("\n".join(matched_texts)).text[0]
-
-    return Pipeline
-
-
-Pipeline = simple_pipeline_cls
+from .simple_pipeline import Pipeline
 
 
 class TestPromptConfig:
-    def test_export_prompt_config(self, simple_pipeline_cls):
+    def test_export_prompt_config(self):
         """Test if the prompt config is exported correctly"""
-        pipeline = simple_pipeline_cls()
+        pipeline = Pipeline()
         config_dict = export_pipeline_to_config(pipeline)
         config = list(config_dict.values())[0]
 
@@ -78,9 +26,42 @@ class TestPromptConfig:
 
 
 class TestPromptUI:
-    def test_uigeneration(self, simple_pipeline_cls):
+    def test_uigeneration(self):
         """Test if the gradio UI is exposed without any problem"""
-        pipeline = simple_pipeline_cls()
+        pipeline = Pipeline()
         config = export_pipeline_to_config(pipeline)
 
         build_from_dict(config)
+
+
+class TestExport:
+    def test_export(self, tmp_path):
+        """Test if the export functionality works without error"""
+        from pathlib import Path
+
+        import yaml
+        from theflow.storage import storage
+
+        config_path = tmp_path / "config.yaml"
+        pipeline = Pipeline()
+        Path(storage.url(pipeline.config.store_result)).mkdir(
+            parents=True, exist_ok=True
+        )
+
+        config_dict = export_pipeline_to_config(pipeline)
+        pipeline_name = list(config_dict.keys())[0]
+
+        config_dict[pipeline_name]["logs"] = {
+            "sheet1": {
+                "inputs": [{"name": "text", "step": ".", "variable": "text"}],
+                "outputs": [{"name": "answer", "step": "."}],
+            },
+        }
+        with open(config_path, "w") as f:
+            yaml.safe_dump(config_dict, f)
+
+        export_from_dict(
+            config=str(config_path),
+            pipeline=pipeline_name,
+            output_path=str(tmp_path / "exported.xlsx"),
+        )